Artificial Intelligence and Corporate Green Transition--Evidence From China's AI Innovation Development Pilot Zones

Artificial Intelligence and Corporate Green Transition--Evidence From China's AI Innovation Development Pilot Zones | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Artificial Intelligence and Corporate Green Transition--Evidence From China's AI Innovation Development Pilot Zones Wenrui Ma This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8993678/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 12 You are reading this latest preprint version Abstract In the context of the synergistic promotion of the digital economy and green transition, the impact of AI development on enterprises' green transition is receiving increasing attention. Based on data on A-share listed companies in Shanghai and Shenzhen from 2010 to 2023, this paper constructs a quasi-natural experiment using the AI pilot test zone policy and employs a double-difference model to systematically identify the impact of AI policy on enterprises' green transition behaviour. It is found that the construction of AI pilot zone significantly promotes the improvement of enterprises' green transition level, and this conclusion still holds under multiple robustness tests. Mechanism analyses show that AI policies primarily promote corporate green transition through multiple pathways, including increasing corporate green investment, strengthening credit availability, improving the efficiency of green innovation, and reducing agency costs. The heterogeneity test based on the TOE analysis framework further reveals that the above policy effects are more significant in firms with a higher degree of digital transformation, a sufficient reserve of high-tech talent, better internal governance and better ESG performance; meanwhile, the policy promotion effect is further enhanced in regions with higher intensity of intellectual property rights (IPR) protection systems and environmental regulations. This paper reveals the mechanisms by which AI policies promote green transition at the micro-firm level, providing new empirical evidence for understanding AI-driven green development. Business and commerce/Business and management Social science/Business and management Earth and environmental sciences/Environmental social sciences artificial intelligence corporate green transition green investment credit access green innovation agency cost Figures Figure 1 Figure 2 1. Introduction With the global rise of the green economy, the call for green industrial development has grown increasingly urgent. As key players in industrial sectors and primary contributors to carbon emissions and pollution control, enterprises' green development has become a vital driving force for optimising industrial structures, enhancing productivity levels, and achieving high-quality economic growth. In China, all sectors of society have come to recognise the imperative to shift away from the previous extensive development model, characterised by 'high input, high consumption, and high pollution'. This requires driving the green transition of the economy to achieve high-quality development. In 2022, the report to the 20th National Congress of the Communist Party of China emphasized the need to ‘promote the high-end, intelligent, and green upgrading of traditional industries, and build a green, low-carbon, circular economic system’. Furthermore, the 2025 ‘Proposal of the Central Committee of the Communist Party of China on Formulating the 15th Five-Year Plan for National Economic and Social Development' proposed 'promoting technological transformation and upgrading, advancing the digital and intelligent transformation of manufacturing, developing smart manufacturing, green manufacturing, and service-oriented manufacturing, and accelerating the transformation of industrial models and enterprise organisational forms.' Guided by national strategy, China's enterprises have now entered a new phase of systematic advancement and scaled development in green growth. Regarding the green manufacturing system, by 2025, China had cumulatively cultivated 6,430 green factories, with their output accounting for over 20% of total manufacturing output. In green innovation, the State Intellectual Property Office's Green and Low-Carbon Patent Statistical Analysis Report (2025) indicates that China's low-carbon patent inventions totaled 120,000 in 2024, accounting for nearly 50% of global growth. Chinese enterprises dominate green innovation, accounting for over 70% of green patent applications, and are driving vigorous green and low-carbon technological innovation. Furthermore, in corporate sustainability disclosure, 2,532 A-share listed companies published ESG-related reports in 2025, achieving a disclosure rate of approximately 46.58% and a cumulative net increase of 20% over five years. However, constrained by factors such as imperfect corporate management systems, difficulties in penetrating the green technology market, a lack of clear green standards, and obstacles to industrial-chain emission reductions, the current advancement and effectiveness of corporate green transition are hindered. Effectively promoting corporate green transition not only enhances sustainable development levels but is also pivotal to driving national industrial transformation and comprehensive green development across the economy and society. In recent years, with the advent of a new wave of technological revolution and the deepening of industrial transformation trends, artificial intelligence has become a vital engine for the deep integration of traditional production factors and data elements (Zhang & Zhang, 2020 ). As an inherent component of modern industrial systems, artificial intelligence is a generalised science encompassing methods, theories, and applied systems designed to simulate and augment human intelligence. The technologies (products) it generates are regarded as pivotal drivers of the ‘Fourth Industrial Revolution’. Currently, disruptive technologies centred on AI continue to give rise to new industries, new models and new kinetic energy. China has taken several measures to promote AI applications and technological innovation and has released the New Generation Artificial Intelligence Development Plan in 2017. On 21 May 2019, the Ministry of Industry and Information Technology (MIIT) and the Shanghai Municipal People's Government (Municipal People's Government) jointly inaugurated the first pilot zone in Shanghai, the "Shanghai ( Pudong New Area) Pilot Zone for Artificial Intelligence Innovation and Application", marking the official landing of the National Pilot Zone for Artificial Intelligence Innovation and Application Plan (hereinafter collectively referred to as the AI Policy). Concurrently, in October of the same year, Shenzhen and Qingdao joined Shanghai as the first batch of pilot cities. In February 2021, the second cohort of pilot cities included Beijing, Tianjin, Hangzhou, Guangzhou, and Chengdu. By September 2022, Nanjing, Wuhan, and Changsha were approved as pilot cities. As of February 2026, eleven cities nationwide had been authorised to establish pilot zones. Functioning as core platforms for incubating and scaling intelligent solutions, these AI pilot zones foster regional resource coordination and integration through a dual-track approach of policy support and market-driven initiatives. This facilitates the development of technology validation and application scenarios(Mijit et al., 2025a ). Moreover, within the pilot zone, the introduction of artificial intelligence technology enables enterprises to significantly enhance resource utilisation efficiency, reduce environmental pollution, improve production sustainability, and overcome challenges in the green transition. Taking the AI pilot zone policy as the research entry point, this paper constructs a quasi-natural experiment to identify the causal impact of AI development on the green transition behaviour of enterprises, which expands the existing research in terms of research perspectives, identification strategies and mechanism analysis. Firstly, this paper systematically examines the impact of AI on enterprise green transition from a policy perspective, breaking through the limitations of existing studies that mainly analyse it from the perspectives of digital technology application or enterprise digital transformation. Second, this paper leverages the exogenous policy shock from the establishment of AI pilot zones and employs a multi-period double-difference model to mitigate potential endogeneity, providing empirical evidence with greater causal explanatory power for identifying the promotion of AI for green transition. Again, this paper constructs a multiple transmission mechanism of financing constraint alleviation, green innovation efficiency enhancement and agency cost reduction, revealing the intrinsic role path of AI in promoting green transition of enterprises. Finally, using the Technology-Organisation-Environment (TOE) analysis framework, this paper systematically examines the heterogeneity of policy effects across the dimensions of enterprise resource endowment and institutional environment, deepening understanding of the interaction mechanism between AI and green transition. 2. Literature Review At a time when global environmental problems are becoming increasingly severe, green transition of enterprises has become a core issue of environmental governance in China and even globally. In this context, how to fully leverage the subversive and revolutionary technological advantages of artificial intelligence to empower enterprises' green transition has become a subject of growing concern in the academic community. In the green transition of enterprises, the existing literature primarily focuses on three aspects: conceptual definitions, measurement methods, and factors affecting it. In terms of defining the concept of green transition, the Chinese Academy of Social Sciences (CASS) suggests that green transition is a process in which industries move towards 'intensive use of energy and resources, reduction of pollutant emissions, reduction of environmental impacts, improvement of labour productivity, and enhancement of the capacity for sustainable development’ (Li et al., 2011). In this process, the green transition of enterprises is more evident in the shift in production mode from high energy consumption and high emissions to low energy consumption and low emissions (Kemp & Never, 2017). This paper combines the views of other scholars to further deepen the connotation of green transition, which is understood as a development model that requires enterprises to consider both economic and environmental performance, and to achieve ecological improvement and green economic and social development through industrial upgrading as well as technological innovation (Meng et al., 2024). In measuring enterprises' green transition, existing studies primarily focus on composite index methods, quantitative text analysis methods, single-indicator methods, and DEA methods. For instance, numerous scholars employ the Analytic Hierarchy Process (AHP), linear weighting methods (Z. Chen et al., 2021), entropy weighting(Zhang, 2024), or principal component analysis (PCA) combined with factor analysis (Shi et al., 2020) to construct comprehensive index measurement frameworks. These incorporate multiple factors, such as economic transformation, technological transformation, and the energy transition (Zhai et al., 2022), into assessments of industrial and regional green transitions; Some scholars, adopting a corporate strategic orientation perspective, extract keyword frequencies such as 'carbon', 'carbon dioxide emissions', and 'greenhouse gas (GHG) emissions' from corporate annual reports and social responsibility reports. These are then scored on a 0-5 scale to quantify corporate green transition behaviour (Zhou et al., 2020); Additionally, researchers have employed proxy variables such as reductions in corporate carbon emissions (Li et al., 2013), the proportion of revenue from polluting industries relative to the top five revenue streams (Yang & Chi, 2023), and investment in green transition projects (Wang et al., 2021)for direct measurement. The aforementioned approaches neglect the social benefits associated with resource and environmental factors(Chen & Golley, 2014; Tian & Lin, 2017) and exhibit issues such as highly subjective indicator selection and limited measurement dimensions, thereby failing to reflect the substantive level of corporate green transition. Consequently, numerous scholars have built upon traditional DEA models, employing super-efficient SBM and GML models (Oh, 2010; Tone, 2001) to further incorporate energy consumption and environmental pollution-related conceptual indicators (Li et al., 2013) into total factor productivity (TFP) calculations. These approaches also introduce two categories of variables: undesirable outputs and energy inputs(Zhao et al., 2024), thereby constructing a green total factor productivity (GTFP) measurement model. As an indicator better suited to evaluating levels of corporate green transition, green TFP—through objective weighting, simultaneous handling of multiple inputs and outputs, and efficiency diagnostics—enables more systematic and precise measurement of relative efficiency and improvement pathways. Consequently, it has garnered increasing academic attention(Du & Li, 2019; Jiakui et al., 2023; Zhao et al., 2022). Numerous studies have employed it as an explanatory variable to investigate the impact of smart manufacturing(Cai et al., 2025), fintech(Li & Wang, 2025), digital transformation, the digital economy (Chen et al., 2025), and environmental policies (Liu et al., 2024; Shaopeng Zhang et al., 2024). To avoid misjudging firms' levels of green transition, this paper uses green total factor productivity as a proxy. Regarding the factors influencing the green transition, most scholars have examined the inhibitory or promotional effects of various elements on corporate green transition from three perspectives: policy influence, environmental regulation, and intelligentisation. Firstly, with the state's increasing focus on the green transition of development models, a series of policy interventions have been shown to exert significant incentive effects on corporate green transitions. These measures extend beyond direct incentives such as green credit (Yu & Zhou, 2023), environmental R&D subsidies (Tang & Yang, 2022), and tax and fee reductions (Huseynov & Klamm, 2012), but also extend to institutional innovations such as green finance (G. J. Chen et al., 2021), carbon emissions trading (Li et al., 2025), and industrial internet pilot programmes (Yu & Chen, 2023). Secondly, corporate willingness to pursue green transition is often constrained by profit maximisation considerations (Maghyereh et al., 2025). Against this backdrop, environmental regulations emerge as a driving factor for corporate green transition. For instance, mandatory social responsibility disclosures and clean production industry technical standards strengthen oversight of corporate legitimacy motives, impose stricter requirements on pollution emission intensity, and thereby advance corporate green transition(Wang & Ning, 2020). However, in the long term, environmental regulations may crowd out corporate R&D investment, thereby suppressing innovation capabilities and hindering green transition (Yuan & Xiang, 2018). Overall, environmental regulations exhibit positive, negative, and non-linear relationships with corporate green transition (Liu et al., 2022). Thirdly, from an intelligent perspective, digital and intelligent technologies enhance corporate green awareness and incentivise green innovation by integrating digital and intelligent applications, thereby empowering green transition (Kuang et al., 2024). Additionally, scholars examining industrial robot imports note that while such robots boost production capacity, they also enable more sustainable emission-reduction pathways through green manufacturing, thereby advancing the integration of economic and environmental benefits (Torrent‐Sellens et al., 2025). Furthermore, specific climate risks, such as water scarcity and resource depletion, significantly compel companies to intensify their research and development efforts. This drives the development of green, intelligent supply chains, thereby compelling enterprises to undergo green transition (Fang, 2024). The National Pilot Zone for Artificial Intelligence Innovation and Application is a demonstration zone approved by China's Ministry of Industry and Information Technology to promote the deep integration of artificial intelligence with the real economy. The construction of the pilot zone is driven by the opening of application scenarios, creating a curator of AI innovation and an industrial highland through core technology research, the layout of arithmetic infrastructure, and the cultivation of an industrial ecosystem. Since 2019, Shanghai, Shenzhen, and Jinan-Qingdao have been designated as the first batch of national AI innovation and application pilot zones. As of 2025, 11 cities across the country have been included in the pilot zones, covering Shanghai, Shenzhen, Jinan-Qingdao, Beijing, Tianjin, Hangzhou, Guangzhou, Chengdu, Nanjing, Wuhan, and Changsha. Based on their respective advantages, the zones focus on intelligent manufacturing, smart cities, healthcare, autonomous driving, and other areas, promoting the deep integration of AI and the real economy. Based on regional resource endowments, the pilot regions have formed a differentiated development pattern, as shown in Table 1. Table 1: Overview of China's National AI Innovation Development Pilot Zones Batch Approval Date Pilot Zone Name Distinctive Application Fields and Development Priorities First Batch May/October 2019 Shanghai (Pudong) Focuses on AI chips (Zhangjiang), smart finance (Lujiazui), intelligent manufacturing and smart healthcare. Promoting deep integration of AI algorithms, chips and application scenarios. First Batch May/October 2019 Shenzhen Leveraging its robust electronics industry, prioritising AI chips, smart hardware (particularly smart terminals), smart finance and AI-assisted urban governance. First Batch May/October 2019 Jinan-Qingdao This dual-core pilot zone sees Jinan focusing on smart healthcare and intelligent software; Qingdao, leveraging enterprises like Haier, concentrates on industrial internet and smart home systems while developing smart marine technologies. Second batch February 2021 Beijing Capitalizing on its top-tier scientific talent, prioritises algorithm innovation and security, smart cities (Urban Brain), smart government services, and intelligent connected vehicles. Second batch February 2021 Tianjin (Binhai) Leveraging its port and manufacturing strengths, it prioritises the development of smart ports (Tianjin Port), intelligent manufacturing, and IT application innovation. Second batch February 2021 Hangzhou Leveraging its digital economy strengths (e.g., Alibaba), prioritising development of City Brain (origin), smart retail/e-commerce, smart finance (FinTech), and smart healthcare. Second batch February 2021 Guangzhou Capitalizing on its commercial and automotive industry foundations, focusing on intelligent connected vehicles, smart logistics, smart cities (transport governance), and smart healthcare. Second batch February 2021 Chengdu Leveraging its status as a western hub for healthcare and cultural tourism, prioritising smart healthcare (Huaxi Hospital, etc.), smart cultural tourism, smart finance, and smart agriculture. Third batch September 2022 Nanjing Capitalising on its robust software industry foundation, focusing on intelligent software, smart chips, intelligent manufacturing, and smart grids.。 Third batch September 2022 Wuhan Leveraging the ‘China Optics Valley’ and scientific-educational resources, prioritising development in intelligent manufacturing (particularly intelligent optoelectronics), intelligent remote sensing, intelligent connected vehicles, and smart healthcare. Third batch September 2022 Changsha Capitalising on its distinctive industries, focusing on intelligent construction machinery (e.g., Sany Heavy Industry), intelligent connected vehicles, and smart cultural/creative industries/media (Malan Mountain). Existing research primarily examines the economic effects of artificial intelligence pilot zones from three perspectives: labour reallocation, corporate innovation, and corporate green development. First, artificial intelligence may reshape skill requirements and occupational opportunities, raising concerns about large-scale technological unemployment(Frank et al., 2019). Against this backdrop, pilot zone development has significantly promoted labour reallocation from services to manufacturing by raising entry barriers for low-skilled workers into high-end productive services while reducing skill requirements in manufacturing, thereby enhancing labour resource allocation efficiency (Wang et al., 2024). The establishment of pilot zones also supports employment at both the urban and enterprise levels through industrial agglomeration, expanded corporate markets, and job creation effects (Shen & Zhang, 2024), thereby promoting high-quality full employment. Second, corporate innovation is often constrained by factors such as uncertain returns (K. Wang et al., 2025), policy instability (Barker Iii & Duhaime, 1997), and the innovation performance expectation gap (Manzaneque et al., 2020). The establishment of pilot zones, through stable policy mechanisms, enhances corporate profitability, financing capacity, and technological capabilities, thereby supporting both overall innovation and breakthrough innovation (Fan et al., 2021). Additionally, scholars note that artificial intelligence pilot zones help enterprises secure greater government digital subsidies, reduce profit uncertainty, and attract regional digital talent and capital, thereby creating favourable conditions for corporate innovation and digital transformation (Han et al., 2025; Yuxin & Zhengchu, 2025). Furthermore, Razia found that the development of pilot zones significantly promotes the adoption of artificial intelligence within enterprises(Mijit et al., 2025b). Enhancing technological integration in turn elvates green innovation performance, collectively fostering sustainable innovation growth. Third, green development is a core tenet of the new development philosophy. From an environmental-constraints perspective, Jin conducted an empirical analysis using Chinese provincial panel data, which indicates that establishing AI pilot zones significantly reduced provincial dependence on natural resources (Chang & Yongjian, 2025). This was achieved by enhancing corporate technological efficiency and adjusting industrial structures, thereby further realising corporate green development. Lin, adopting an energy structure optimisation perspective, further demonstrates that AI pilot zone policies markedly enhance energy utilisation efficiency in pilot cities (Lin & Yang, 2025). This effect is particularly pronounced where urban environmental regulation, economic development, and infrastructure construction are already at leading levels, ultimately advancing the realisation of corporate green development objectives. Existing studies have mainly focused on the impact of policy environment, environmental regulation and digital technology on the green transition of enterprises; in the field of artificial intelligence, studies have mainly focused on its role in the productivity, technological innovation and governance structure of enterprises, but there is a lack of systematic analyses of the causal impacts and multiple mechanisms of artificial intelligence in promoting the green transition of enterprises. In addition, the synergistic role of enterprises' internal resource endowment and external institutional environment in green transition has not been fully examined. To address the above shortcomings, this paper takes the AI pilot zone policy as a quasi-natural experiment, systematically identifies the causal impact of AI on enterprise green transition, and analyses the role paths from the multi-dimensional perspective of financing constraints, green innovation efficiency, and governance mechanisms, and at the same time examines the heterogeneity between enterprises and the institutional environment based on the TOE framework, so as to enrich the existing research in the level of policy identification and mechanism integration. 3. Theoretical Analysis and Research Hypotheses Based on stakeholder theory, signaling theory, social learning theory and principal-agent theory, this paper argues that the AI Innovation and Application Pilot Zone policy can promote green transition of enterprises by enhancing the level of green investment, credit availability, green innovation efficiency as well as reducing the agency costs. Firstly, from the perspective of corporate green investment: on the one hand, grounded in stakeholder theory, artificial intelligence as a novel general-purpose technology can optimise production processes and enhance productivity, thereby reducing costs, increasing profits, and satisfying management's short-term performance incentives, thus providing endogenous funding for green investment. Simultaneously, AI continuously generates high-frequency, multidimensional operational and energy-consumption data. Green investors, as stakeholders, can utilise this data to identify a firm's green transition potential, thereby increasing exogenous green investment support. Furthermore, AI can directly empower green investment decision-making by predicting energy demand and evaluating project returns alongside environmental benefits, thereby reducing investment uncertainty and strengthening management's willingness to invest. On the other hand, increased green investment will significantly drive corporate green transition. From a financing perspective, green investment directly funds corporate green transition projects, alleviating financing constraints (Zhang & Sun, 2023 ). From a governance perspective, green investment often entails specific environmental performance targets or market preferences (Yan et al., 2021 ), guiding enterprises to optimise product portfolios by increasing R&D investment and production share of green products (Chen & Ma, 2021 ; Zhang et al., 2020 ), and establish green supply chains (Adnan et al., 2025 ). These actions elevate corporate green transition levels on the supply side, meeting market expectations. In addition, based on the perspective of external supervision, as the scale of green investment expands, external investors such as green funds and banks will be involved in corporate management decisions (Kim & Yoon, 2023 ), which not only strengthens the supervision of corporate behaviours, but also significantly inhibits ‘greenwashing’ (Qian et al., 2025 ) and opportunistic tendencies (Pang et al., 2025 ), ensuring that funds really flow to green transition activities, encouraging shareholder activism (Zhu et al., 2025 )to oversee firms to strengthen green supply chain management and improve green innovation to realise the green transition (A. Wang et al., 2025 ; Zhu et al., 2023 ). Second, from the perspective of enterprise credit availability, on the one hand, in traditional financial practices, enterprises mainly rely on establishing close ties with banks to obtain credit (Berlin & Mester, 1998 ), and enterprise credit availability is subject to the double constraints of information asymmetry between banks and enterprises and the transaction costs of loans (Chong et al., 2013 ), which results in the green transition activities of enterprises being hindered by the problems of 'difficult financing’ and ‘expensive financing’. To address the above problems, based on signaling theory, AI technology can help banks to effectively capture enterprise internal operation information by promoting the intelligence and dataisation of enterprise production process, so that enterprises can leave enough ‘digital footprints’, such as payment records and logistic information (Babina et al., 2024 ), and thus help banks to effectively capture enterprise internal operation information (Babina et al., 2024 ), which can help banks to effectively capture enterprise internal operation information (Agrawal et al., 2019 ; Frank et al., 2019 ; Hilb, 2020 ), and reduce the information asymmetry between banks and enterprises; at the same time, AI can transform non-standardised data already available to enterprises into visual information (Agrawal et al., 2018 ; Babina et al., 2024 ), simplifying the bank review process, compressing the transaction cost and approval time (Chen et al., 2013 ), reducing the transaction cost of enterprise loans, and thus enhancing credit availability. On the other hand, increased credit availability substantially contributes to the green transition. Firstly, improved financing conditions and extended debt repayment cycles provide enterprises with long-term, stable funding for energy-saving upgrades and green innovation (Herrera & Minetti, 2007 ; Shujing Zhang et al., 2024 ). Secondly, sustained and enhanced credit accessibility continuously incentivises enterprises to optimise internal operations and environmental disclosure, deepening green practices (Flammer, 2021 ). Finally, grounded in legitimacy theory, an implicit social contract exists between enterprises and society (Mäkelä & Näsi, 2010 ; Sacconi, 2006 ). Enhanced credit accessibility attracts greater stakeholder scrutiny, generating contractual pressure (De Haas et al., 2025 ). To safeguard their reputation and reduce capital costs, firms have incentives to advance green transition in alignment with societal value norms, thereby enhancing legitimacy(Chen et al., 2009 ; Shahzad et al., 2024 ). Furthermore, from the perspective of corporate green innovation efficiency, artificial intelligence facilitates green transition by elevating the efficacy of green innovation. On the one hand, grounded in social learning theory, corporate green innovation relies on inventors' capacity to absorb external information and engage in mutual intellectual stimulation (De Faria et al., 2010 ). Leveraging technological advantages such as machine learning, intelligent perception, and big data analytics, artificial intelligence can proactively process and deeply mine data assets to generate innovative knowledge insights (Jia et al., 2024 ). Consequently, inventors can establish novel social networks characterised by heightened innovative thinking through human-machine collaboration (Babina et al., 2024 ). Leveraging the broader knowledge spillover effects generated by such networks (Raisch & Fomina, 2025 ), this approach enhances cross-firm and cross-industry information exchange, thereby facilitating technological convergence (Mijit et al., 2025b ), significantly enhancing corporate efficiency in green innovation product development and technology commercialisation rates (Lee et al., 2024 ). Moreover, artificial intelligence can elevate corporate demand for highly skilled labour through "machine substitution for human labour" (Acemoglu & Restrepo, 2018 ), thereby reducing barriers to entry and trial-and-error costs in green innovation, achieving an overall efficiency leap. Conversely, enhanced green innovation efficiency continuously elevates corporate green transition levels. First, grounded in efficiency theory, the synergistic increase in both the quantity and quality of corporate green innovations cultivates technological advantages in green production. This enhances resource utilisation efficiency at the supply end, reduces pollution during production processes, and ultimately guides the green transition of corporate production workflows. Secondly, on the demand side, green product innovation responds to market trends towards low-carbon consumption. Through differentiation strategies, it enhances brand value and reputation (Yang & Chi, 2023 ), thereby generating new business opportunities and deepening firms' willingness and capacity for green transition. Finally, from the perspective of corporate agency costs: firstly, based on principal-agent theory, these costs arise from the separation of ownership and management rights within firms, comprising principal monitoring costs, agent collateral costs, and residual losses (Jensen, 1986 ). Artificial intelligence technology can mitigate these agency costs by enhancing information transparency through sampling analysis and real-time alerts of vast internal operational data and external public sentiment. Furthermore, AI is entering a new phase of continuous learning and self-improvement, with its behavioural patterns shifting from simple human mimicry toward autonomous decision-making capabilities (Vanneste & Puranam, 2024 ). Aligning AI's autonomous decision-making objectives with long-term corporate value not only leverages AI's "rational agent" advantage (Humberd & Latham, 2025 ), reducing cost stickiness caused by agent self-interest and short-termism (Chen et al., 2024 ), optimising corporate resource allocation. Moreover, more sophisticated predictive models can minimise residual losses stemming from decision biases within principal-agent relationships (Humberd & Latham, 2025 ). Conversely, reduced agency costs further empower corporate green transitions. Firstly, improved incentive structures align management interests with long-term corporate development. To build sustainable competitive advantages, management demonstrates greater willingness to advance green transition(Li et al., 2024 ; Martin et al., 2019 ). Secondly, firms will persistently rely on high information transparency to maintain low agency costs, making their environmental performance more susceptible to public scrutiny (Geng & Wen, 2025 ). Consequently, green transition gradually becomes a voluntary choice for management. Finally, reduced agency costs signify enhanced resource utilisation efficiency (Aktas et al., 2019 ), directing greater resources towards corporate green governance activities and directly empowering green transition. In summary, this paper proposes the following hypothesis: H1: Artificial intelligence effectively drives corporate green transition. H2: Artificial intelligence facilitates corporate green transition by enhancing green investment levels, credit availability, and the efficiency of green innovation, while reducing agency costs. 4. Research Design 4.1 Model Specification 4.1.1 Baseline Regression Model To examine the impact of artificial intelligence pilot zones on corporate green transition, the following regression model is constructed: 4.1.2 Mediating Effect Model To explore the mechanism of AI pilot zones' role in enterprises' green transition, this paper adopts a mediation effect analysis. Given that the traditional stepwise regression method for estimating mediated effects suffers from endogeneity bias (Jiang, 2022 ), this paper examines the impact of core explanatory variables on mediating variables by replacing the explanatory variables with the mediating variables to test the existence and validity of the mediating path. The mediating effect model is set as follows: where the mediating variables are credit availability (Flex2), corporate green investment (CGI), green R&D efficiency (GRD), green outcome conversion efficiency (GCON), and agency costs (AC). All other variables are consistent with Eq. (1). 4.2 Variable Selection 4.2.1 Dependent Variables Corporate Green transition (CGT): Incorporating environmental pollution into the evaluation framework, the non-radial SBM-ML index (hereafter ‘ML index’) is employed to measure corporate green total factor productivity (Chen et al., 2025 ). This productivity comprises input and output indicators, with inputs consisting of labour, capital, and energy inputs, while outputs encompass desired and undesired outputs. This paper employs the following methodology to measure input and output indicators: (1) Input factors: Labour input is proxied by the number of enterprise employees; capital input is proxied by net fixed assets; energy input is proxied by the enterprise's share of industrial electricity consumption in its city's total urban industrial electricity consumption, calculated based on the proportion of enterprise employees relative to the city's total urban employed population. (2) Desired output: enterprise business income as a proxy variable for the desired output of the enterprise. (3) Undesired output: the proportion of employees in the enterprise in the employment of urban workers in the city where it is located is converted to the ‘three industrial wastes', namely, industrial Sulphur dioxide, industrial wastewater and industrial smoke and dust emissions, as a proxy for the enterprise's undesired output. 4.2.2 Explanatory Variables Artificial Intelligence Pilot Zone Policy (inter): as an explanatory variable, the Artificial Intelligence Pilot Zone serves as a core platform for incubating and promoting intelligent solutions and builds AI technology validation and application scenarios through the two-way linkage of policy support and market-driven forces, accumulating replicable and promotable experience for the whole country. In terms of green development, the establishment of AI pilot zones is committed to promoting the in-depth integration of AI technology with green innovation and sustainable development, achieving significant improvement in energy efficiency and optimal allocation of resources through technological innovation, and thus promoting the transformation and development of a low-carbon economy (Guan et al., 2026 ). inter is equal to Treat × Post, in which the number of cities that implement the AI The enterprises matched by the cities implementing the pilot test zone policy are assigned a value of 1 as the experimental group; the remaining enterprises are assigned a value of 0 as the control group. The time dummy variable Post is assigned a value of 1 for the first year of implementing the AI pilot zone policy and beyond, and 0 for the rest of the years (Liu et al., 2026 ). 4.2.3 Mediating Variable Credit Availability (Flex2): Credit availability is a core indicator of an enterprise's ability to obtain bank credit resources, and this paper follows the established research paradigm, using the ratio of interest-bearing liabilities to total assets to measure it (Flex2 = interest-bearing liabilities/total assets). Among them, the accounting calibre of interest-bearing liabilities is defined as the sum of short-term borrowings, interest payable, non-current liabilities due within one year, long-term borrowings and bonds payable. The higher this indicator's value, the stronger the enterprise's ability to obtain credit. As a key link between the financial market and business operations, the level of credit availability has a profound impact on the efficiency of the enterprise's capital turnover and the potential for investment and development. Especially during the monetary policy adjustment cycle, the high sensitivity of credit availability to changes in the financial environment makes it a key indicator of shifts in corporate financing constraints. Corporate Green Investment (CGI): Green investment can be measured by the amount of investment in environmental protection, i.e., the amount of investment in environmental protection projects in progress (Tian et al., 2024 ) (e.g., desulphurisation and denitrification, wastewater treatment, exhaust gas treatment, clean innovation, procurement of environmental protection equipment, etc.). The measurement of this variable is primarily based on studies of CSR or green investment. By aggregating the various expenditures related to pollution prevention, ecological management and green production in the financial statements of enterprises to derive the total annual green investment, and then dividing it by the total assets at the end of the period for standardisation to get the green investment, it can more accurately measure the strength and scale of the enterprises' investment in green transition. In this paper, we refer to Jiang (Jiang et al., 2025 ) and Huang (Huang & Lei, 2021 ), and measure the intensity of green investment by the ratio of environmental protection construction in progress/end-of-period total assets of enterprises. Green science and technology R&D efficiency (GRD) and green achievement conversion efficiency (GCON): drawing on the measurement idea of Bi et al. (2016) (Bi et al., 2016), in the stage of green science and technology R&D, the number of enterprise R&D personnel and R&D expenditure are selected as the initial inputs, and the number of enterprise green patent applications and the number of enterprise green patent authorisations are taken as the intermediate output indicators. In the green results application stage, the intermediate outputs are the number of enterprise green patent applications and the number of enterprise green patent authorisations, while the sales revenue, enterprise pollution emission index and enterprise energy consumption index are used as the final output indicators. The efficiencies of both stages are calculated using the advanced DEA-SBM model. Agency cost (AC): corporate agency cost is a kind of expense due to the separation of corporate ownership and operation, the higher the agency cost, the more serious the agency conflict between management and shareholders. In this paper, the following indicators are selected as proxy variables for corporate agency costs: (1) operating expense ratio (AC1); (2) management expense ratio (AC2); (3) asset turnover ratio (AC3). 4.2.4 Control Variables With reference to previous studies, the following variables that may affect the green transition of enterprises are selected: enterprise size (SIZE): expressed as the logarithmic value of the enterprise's total assets at the end of the year; shareholding ratio of the first largest shareholder (TOP): expressed as the shareholding share of the first largest shareholder divided by the company's total share capital; gearing ratio (LEV): expressed as the enterprise's total liabilities at the end of the year divided by the total assets at the end of the year; profitability of total assets (ROA). Net profit divided by total assets; CASH: Net cash flow from operating activities divided by total assets; DUAL: 1 if the chairman of the board of directors is the same person as the managing director, 0 otherwise; MBOARD, measured by the number of board members. Gross regional product (GDP): Ln(city GDP). Regional industrial upgrading (INDUS) is defined as the value added of the regional tertiary industry divided by the value added of the secondary industry. 4.3 Data source and processing This paper uses A-share listed companies from 2010 to 2023 as the research sample, removes missing observations, and shrinks the data by 1%, yielding 36,449 observations. The main explanatory variables, the list of AI pilot test zones, are from the Ministry of Industry and Information Technology (MIIT), the data at the enterprise level are from the WIND and the Cathay Pacific CSMAR database, the data at the city level are from the City Statistical Yearbook, and the data processing and regression analyses are done with Stata 18.0. The descriptive statistics of the variables are detailed in Table 4.1 . Table 4.1 Descriptive Statistics VarName Obs Mean SD Min Median Max CGT 36449 1.0281 0.1079 0.8000 1.0419 1.2553 inter 36449 0.1418 0.3488 0.0000 0.0000 1.0000 ROA 36449 0.0403 0.2381 -7.2851 0.0375 29.6619 SIZE 36449 22.2593 1.3420 13.0760 22.0704 28.6969 LEV 36449 0.4396 0.4366 0.0080 0.4205 31.4667 TOP 36449 0.3386 0.1499 0.0029 0.3150 0.8999 MBOARD 36449 2.1269 0.1975 0.6931 2.1972 2.9444 CASH 36449 0.2306 0.5209 -25.0434 0.1452 13.3110 DUAL 36449 0.2741 0.4461 0.0000 0.0000 1.0000 GDP 36449 10.5984 0.7960 6.2400 10.6807 11.8180 INDUS 36449 2.4792 0.1474 2.1323 2.4502 2.8461 5. Empirical Findings 5.1 Baseline Regression Results Table 5.1 reports the results of the baseline regression of the AI pilot zone policy on firms' green transition, where column (1) is the regression model without control variables, and columns (2) to (5) are the regression setups with no fixed year and individual, fixed effects of year only, fixed effects of individual only, and fixed effects of year and individual in both directions, respectively. The results show that the regression results in columns (1) to (5) are negative and significant, and the regression coefficient of the core explanatory variable AI pilot test area (inter) in column (5) is 0.0021 and is significant at the 1% level. This indicates that all the AI pilot zone policies significantly promote green transition of enterprises. Table 5.1 Baseline Regression Results (1) (2) (3) (4) (5) CGT CGT CGT CGT CGT inter 0.0021*** 0.0877*** 0.0020*** 0.0257*** 0.0021*** (5.183) (61.398) (6.278) (48.000) (5.107) ROA -0.0124*** 0.0003 -0.0010* 0.0002 (-6.380) (0.844) (-1.667) (0.427) SIZE 0.0144*** 0.0001 0.0031*** -0.0000 (39.293) (1.392) (10.576) (-0.161) LEV -0.0078*** 0.0002 0.0013*** 0.0001 (-7.190) (1.036) (3.586) (0.314) TOP -0.0883*** -0.0015** -0.0298*** -0.0009 (-28.481) (-2.352) (-13.792) (-0.632) MBOARD -0.0612*** -0.0012** -0.0074*** -0.0016* (-25.290) (-2.471) (-5.681) (-1.733) CASH 0.0063*** 0.0001 0.0001 -0.0001 (7.127) (0.281) (0.392) (-0.413) DUAL 0.0031*** -0.0002 -0.0022*** -0.0006* (2.970) (-1.098) (-4.700) (-1.887) GDP 0.0446*** 0.0000 0.2769*** -0.0022 (74.537) (0.118) (276.869) (-1.093) INDUS 0.0796*** 0.0006 0.2401*** -0.0052 (23.587) (0.862) (42.123) (-0.999) _cons 0.8204*** 0.1880*** 0.8196*** -2.5491*** 0.8589*** (1764.204) (14.723) (303.194) (-283.374) (33.491) FE Yes No No Yes Yes Year Yes No Yes No Yes N 36449 36449 36449 36449 36449 R 2 0.9778 0.3630 0.9741 0.9533 0.9778 Note: ***, ** and * denote significance levels of 1%, 5% and 10% respectively. The same applies to the table below. 5.2 Parallel Trends Test The validity of the double difference model relies on the fulfilment of the parallel trend assumption, i.e. the outcome variables of the experimental group and the control group should show similar trends before the policy intervention, otherwise it may lead to biased estimation of the policy effect. In order to verify this assumption, this paper takes the year before the policy implementation as the base period and constructs a multi-period double difference model to test the trend consistency of the green transition of enterprises, and the results are shown in Figure 1. As can be seen from the figure, at each observation point before the policy intervention (pre_5 to pre_2), the coefficients of enterprise green transition of the experimental group and the control group fluctuate around the value of 0 in a small way, and the corresponding confidence intervals contain 0, which indicates that there is no significant difference in the trend of change of the green transition of the enterprises in the two groups, and that the hypothesis of parallel trend is valid. And after the implementation of the policy (pos_1 and after), the enterprise green transition coefficient changes significantly, indicating that the policy has a significant promotion effect on enterprise green transition. 5.3 Placebo Test To control for potential omitted variables, this paper tests whether the policy effect is due to random factors using a placebo test. Specifically, a randomised treatment group is constructed using a multi-temporal double-difference method: a sample of the same size as the original treatment group is randomly selected, and a treatment group variable and a policy time-point variable are randomly generated. 500 random policy shocks are simulated for the whole sample, and the resulting output, shown in Figure 2, indicates that the coefficients of the random interaction term are centered around 0 and generally smaller than the true estimate of 0.0021. This suggests that the policy effect is significantly weakened after randomisation, supporting the robustness of this paper's conclusions. 5.4 Propensity Score Matching and Robustness Testing 5.4.1 Propensity Score Matching Method To mitigate the interference that sample selection bias may bring to the results of the benchmark regression, this paper adopts the Propensity Score Matching (PSM) method to conduct a robustness test to further verify whether the impact of the AI pilot zone policy on the green transition of enterprises is robust. Specifically, a matched sample is constructed using firms' basic characteristics. Using a 1:3 nearest-neighbour matching approach, firms in the pilot zone are re-paired with those excluded. Regression estimation is then conducted on the matched sample. Table 5.2 reports the regression results after propensity score matching. Columns (1) and (2) show regression coefficients of 0.0026 and 0.0028, respectively, for the policy variable inter on corporate green transition, both significant at the 1% level. In summary, after controlling sample selection bias, the AI pilot zones still exert a significant promotional effect on corporate green transition, further validating the robustness of the baseline regression results. Table 5.2 Propensity Score Matching Results (1) (2) CGT CGT inter 0.0026*** 0.0028*** (3.090) (3.227) ROA -0.0024 (-0.876) SIZE 0.0005 (0.723) LEV 0.0017 (0.745) TOP -0.0092** (-2.148) MBOARD -0.0061*** (-2.672) CASH -0.0000 (-0.013) DUAL -0.0000 (-0.046) GDP -0.0070 (-0.447) INDUS 0.0280 (1.224) _cons 0.8205*** 0.8293*** (293.645) (4.581) FE Yes Yes Year Yes Yes N 10436 10436 R 2 0.9662 0.9663 5.4.2 Controlling for the Impact of the Pandemic To further validate the robustness of the baseline regression results, this study treats the COVID-19 pandemic as an exogenous shock that may affect corporate green transition and policy implementation outcomes. The 2020 sample year was therefore excluded to control for the pandemic's interference. The regression results in Column (1) of Table 5.3 show that the regression coefficient for the core explanatory variable is -0.0262, significant at the 5% level. It shows that the role of the AI pilot zone in facilitating companies' green transition remains even after the year of the outbreak is ruled out. 5.4.3 Adding control variables To further control for potential disturbance factors arising from corporate governance structure and business status, this paper introduces the firm's age (AGE) and growth (GROWTH) as additional control variables in the regression. According to the regression results in Table 5.3, paragraph (2), the regression coefficient for the core explanatory variable INTER is -0.0360 and is significant at the 1% level. It shows that the result of the role of AI pilot zone on green transition of enterprises remains robust. 5.4.4 Excluding the effect of non-manufacturing industries In order to further test the robustness of the baseline regression results, this paper selects manufacturing enterprises as the research sample to carry out the retest. The regression results in column (3) of Table 5.3 show that the estimated coefficient of the core explanatory variables is 0.0016 and is significant at the 1 per cent statistical level. This indicates that even after excluding the sample of non-manufacturing enterprises, establishing AI pilot zones still has a significant promotional effect on enterprises' green transition, and the baseline regression conclusions remain robust. Table 5.3 Robustness Testing Exclude 2020 Add control variables Exclude non-manufacturing CGT CGT CGT inter 0.0021*** 0.0021*** 0.0016*** (4.828) (5.115) (3.022) ROA 0.0002 0.0002 0.0005 (0.553) (0.463) (0.716) SIZE -0.0000 -0.0000 -0.0002 (-0.151) (-0.082) (-0.634) LEV 0.0001 0.0001 0.0001 (0.380) (0.363) (0.465) TOP -0.0011 -0.0011 0.0017 (-0.682) (-0.729) (0.868) MBOARD -0.0013 -0.0016* -0.0019 (-1.418) (-1.751) (-1.580) CASH -0.0001 -0.0001 -0.0003 (-0.381) (-0.464) (-1.051) DUAL -0.0007** -0.0006* -0.0009** (-2.102) (-1.924) (-2.224) GDP -0.0028 -0.0021 -0.0024 (-1.372) (-1.084) (-0.938) INDUS -0.0058 -0.0052 -0.0140** (-1.074) (-0.988) (-2.063) GROWTH -0.0000 (-1.557) AGE -0.0002 (-0.415) _cons 0.8664*** 0.8586*** 0.8850*** (32.451) (33.474) (26.507) FE Yes Yes Yes Year Yes Yes Yes N 33287 36449 23664 R 2 0.9789 0.9778 0.9767 5.4.5 Robustness test using dual machine learning Dual machine learning can effectively handle high-dimensional covariates and mitigate endogeneity (Bia et al., 2024), and more accurately estimate the causal impact of the AI pilot zone policy on enterprises' green transition. In this paper, we use a dual machine learning model and apply common algorithms, such as LASSO, random forest, neural network, and gradient boosting, to robustly test the green transition-promoting effect of the AI pilot zone policy. The regression results in columns (1)-(6) of Table 5.4 show that, regardless of the algorithms, the regression coefficients of the green transition of the enterprises by AI pilot zone policy are all significantly positive, further verifying the robustness and reliability of the benchmark regression conclusions. Table 5.4 Comparative results of machine learning methods (1) (2) (3) (4) (5) (6) LASSO Random Forest Neural network Gradient Boosting Elastic Net SVM inter 0.002 *** 0.003 *** 0.003 * 0.005 *** 0.002 *** 0.056 *** (0.000) (0.001) (0.002) (0.001) (0.000) (0.001) Control YES YES YES YES YES YES FE YES YES YES YES YES YES Year YES YES YES YES YES YES N 36449 36449 36449 36449 36449 36449 5.5 Mediating Effect Analysis 5.5.1 Mediating Role of Green Capital Investment (CGI) Column (2) of Table 5.4 indicates that the coefficient for the impact of AI pilot zones on green capital investment is 1.5083, significant at the 10% level. This suggests AI pilot zones can enhance corporate green capital investment. Green capital investment systematically drives corporate green transition through financing, governance, and oversight. On the one hand, green investment directly funds corporate green projects, alleviating financing constraints (Zhang & Sun, 2023). On the other hand, by leveraging specific environmental performance requirements (Yan et al., 2021), it guides enterprises to optimise product portfolios and supply chains, thereby enhancing green supply capabilities(Adnan et al., 2025; Chen & Ma, 2021; Zhang et al., 2020). simultaneously, green investment encourages external investors to participate in corporate governance, curbing greenwashing and similar practices while overseeing enhanced green innovation to ensure transformational efficacy (Kim & Yoon, 2023; Pang et al., 2025; Qian et al., 2025). Therefore, green investment plays an important role in the transmission between AI pilot zones and enterprises' green transition, supporting the existence of mediating effects. 5.5.2 Mediating Role of Credit Accessibility (Flex2) Column (3) of Table 5.4 shows that the coefficient for the impact of AI pilot zones on green investment is 0.0035, which is significant at the 5% level, indicating that AI pilot zones can improve enterprise credit availability. The improvement of enterprise credit availability can provide a stable long-term financial guarantee for enterprise's energy-saving renovation and green innovation (Herrera & Minetti, 2007; Shujing Zhang et al., 2024), and motivate enterprises to optimise their internal operation and improve environmental information disclosure, thus deepening their green practices(Flammer, 2021). Thus, credit availability plays an important role in the transmission between AI pilot zones and corporate green transition, validating the existence of mediating effects. Table 5.5 Mediating Factors of Green Capital Investment and Credit Accessibility (1) (2) (3) CGT CGI Flex2 inter 0.0021*** 1.5083* 0.0035** (5.107) (1.840) (2.219) ROA 0.0002 3.3801*** -0.0912*** (0.427) (3.355) (-13.725) SIZE -0.0000 -3.6547*** 0.0233*** (-0.161) (-10.017) (25.377) LEV 0.0001 12.3043*** 0.5238*** (0.314) (20.852) (137.650) TOP -0.0009 -15.9960*** 0.0209*** (-0.632) (-6.231) (3.555) MBOARD -0.0016* -2.3762 -0.0057 (-1.733) (-1.544) (-1.619) CASH -0.0001 0.2023 -0.0159*** (-0.413) (0.501) (-12.034) DUAL -0.0006* 2.1558*** 0.0003 (-1.887) (3.904) (0.213) GDP -0.0022 -2.4557 -0.0078 (-1.093) (-0.747) (-1.020) INDUS -0.0052 2.0366 -0.0541*** (-0.999) (0.225) (-2.692) _cons 0.8589*** 102.6268** -0.3170*** (33.491) (2.366) (-3.206) FE Yes Yes Yes Year Yes Yes Yes N 36449 29382 31581 R 2 0.9778 0.1051 0.8582 5.5.3 Green Innovation Efficiency Mediators In columns (2) and (3) of Table 5.5, the coefficient of influence of AI pilot zones on the efficiency of green technology R&D is 0.0062, which is significant at the 10% significant level, and the coefficient of influence on the green achievement conversion efficiency is 0.0074, which is significant at the 5% significant level, which suggests that the AI pilot zones can improve the efficiency of green technology R&D and green achievement conversion. According to the efficiency theory, the synergistic improvement of the quantity and quality of green innovation of enterprises can cultivate the advantages of green production technology, promote the greening of the production process by improving resource efficiency and reducing pollution on the supply side, and respond to the trend of low-carbon consumption to enhance brand value and expand market opportunities with differentiated strategies on the demand side, thus systematically deepening the motivation and ability of green transition of enterprises (Yang & Chi, 2023). Therefore, the efficiency of green science and technology R&D and green achievement conversion play important roles in the transmission between the AI pilot zone and the green transition of enterprises, supporting the existence of a mediating effect. Table 5.6 Mediating Factors of Green Technology R&D and Conversion Efficiency (1) (2) (3) CGT GRD GCON inter 0.0021*** 0.0062* 0.0074** (5.107) (1.754) (2.137) ROA 0.0002 0.0017 -0.0036 (0.427) (0.512) (-1.082) SIZE -0.0000 -0.0015 0.0007 (-0.161) (-0.852) (0.381) LEV 0.0001 -0.0016 -0.0006 (0.314) (-0.775) (-0.269) TOP -0.0009 -0.0047 0.0103 (-0.632) (-0.366) (0.809) MBOARD -0.0016* -0.0009 0.0009 (-1.733) (-0.123) (0.117) CASH -0.0001 -0.0010 -0.0003 (-0.413) (-0.529) (-0.141) DUAL -0.0006* 0.0009 -0.0000 (-1.887) (0.322) (-0.003) GDP -0.0022 -0.0038 -0.0114 (-1.093) (-0.226) (-0.682) INDUS -0.0052 0.0261 0.0215 (-0.999) (0.584) (0.483) _cons 0.8589*** 0.2443 0.2737 (33.491) (1.123) (1.262) FE Yes Yes Yes Year Yes Yes Yes N 36449 33714 33826 R 2 0.9778 0.6276 0.6244 5.5.4 Mediating Role of Agency Costs (AC) In columns (2) (3) (4) of Table 5.6, the coefficients of the impact of AI pilot zones on the impact on agency costs are -0.0315 (significant at the 5% level of significance), -0.0237 (significant at the 10% level of significance), and 0.0249 (significant at the 1% level of significance), which suggests that AI pilot zones reduce the agency costs of firms. The reduction of enterprise agency cost empowers green transition through the dual path of governance and resources. First, the highly transparent governance mechanism brought about by the reduction of agency costs can strengthen external scrutiny and supervision of environmental performance and promote the internalisation of green transition within corporate management consciousness (Geng & Wen, 2025). Second, the decrease in agency costs implies that enterprise resource utilisation becomes more efficient, thereby releasing more resources for green governance activities and directly supporting the transformation process (Aktas et al., 2019). Therefore, agency costs play an important role in the transmission between AI pilot zones and corporate green transition, verifying the existence of mediating effects. Table 5.7 Mediating Factor of Agency Costs (1) (2) (3) (4) CGT AC1 AC2 AC3 inter 0.0021*** -0.0315** -0.0237* 0.0249*** (5.107) (-2.511) (-1.927) (3.307) ROA 0.0002 -0.0151 -0.0023 0.0662*** (0.427) (-1.240) (-0.191) (9.481) SIZE -0.0000 -0.0685*** -0.0634*** -0.0624*** (-0.161) (-11.735) (-11.043) (-18.639) LEV 0.0001 0.0483*** 0.0494*** 0.0175** (0.314) (4.038) (4.195) (2.562) TOP -0.0009 -0.0381 -0.0335 -0.0206 (-0.632) (-0.935) (-0.840) (-0.887) MBOARD -0.0016* 0.0124 0.0115 0.0327** (-1.733) (0.502) (0.474) (2.324) CASH -0.0001 -0.1628*** -0.1551*** 0.0193*** (-0.413) (-24.870) (-24.270) (5.225) DUAL -0.0006* 0.0280*** 0.0246*** -0.0095* (-1.887) (3.139) (2.792) (-1.867) GDP -0.0022 -0.0080 -0.0014 0.0436 (-1.093) (-0.432) (-0.079) (1.484) INDUS -0.0052 0.3183*** 0.2213** -0.1346* (-0.999) (3.047) (2.164) (-1.679) _cons 0.8589*** 0.9921*** 0.9736*** 1.8498*** (33.491) (6.014) (6.014) (4.805) FE Yes Yes Yes Yes Year Yes Yes Yes Yes N 36449 31824 32148 32148 R 2 0.9778 0.1732 0.1264 0.7998 6 Heterogeneity Analysis Based on the Technology-Organisation-Environment (TOE) analytical framework, this part empirically examines the impact of AI pilot zones on the green transition of enterprises and focuses on the moderating role of the three types of factors: technology, organisation and environment. 6.1 Technological Capability Technological capability denotes the capacity of corporate management and employees to utilise diverse technological resources. Within the TOE framework, the technological dimension focuses on firms' endogenous technological capabilities. Leveraging mature data infrastructure and employees' efficient technological learning abilities, policy resources can more effectively drive optimisation of green production processes and technological innovation practices. This study predicts that for enterprises with high levels of digital transformation and a high proportion of highly skilled personnel, the policy intensity of AI pilot zones has a stronger promotional effect on green transition. First, this study defines enterprise digital transformation indicators as core metrics to characterise the synergistic interaction between technical capability and pilot zone policies (Wu et al., 2021). Heterogeneity in digital transformation, as shown in Column (2) of Table 6.1, reveals: the coefficient for the impact of AI pilot zone policies on green transition is 0.0017, significant at the 1% level. The coefficient for DT_inter is 0.0005, significant at the 5% level, indicating that for enterprises with higher levels of digital transformation, the AI pilot zone policy has a stronger promotional effect on green transition. Secondly, drawing upon the methodology of Liu (Liu & Zhao, 2020), this study measures the heterogeneity of the proportion of highly skilled personnel within enterprises using the ratio of highly skilled personnel to total employees. The heterogeneity of the proportion of highly skilled personnel, as shown in column (3) of Table 6.1, reveals that the coefficient for the AI pilot zone policy on green transition is 0.0019, significant at the 1% level. The coefficient for SKILL_inter is 0.0001, significant at the 5% level. This indicates that enterprises with higher levels of digital transformation experience a stronger promotional effect of the AI pilot zone policy on green transition. The results in column (3) of Table 6.1 show that the coefficient for the impact of AI pilot zone policies on green transition is 0.0019, significant at the 1% level. The coefficient for SKILL_inter is 0.0028, significant at the 10% level. This indicates that the greater the proportion of highly skilled personnel within an enterprise, the stronger the promotional effect of AI pilot zone policies on green transition. Table 6.1 Technological Capability Heterogeneity (1) digital transformation highly skilled personnel CGT CGT CGT inter 0.0021*** 0.0017*** 0.0019*** (5.107) (3.758) (4.364) ROA 0.0002 0.0002 0.0005 (0.427) (0.437) (0.626) SIZE -0.0000 -0.0000 0.0001 (-0.161) (-0.176) (0.272) LEV 0.0001 0.0001 0.0002 (0.314) (0.302) (0.418) TOP -0.0009 -0.0008 -0.0010 (-0.632) (-0.536) (-0.610) MBOARD -0.0016* -0.0016* -0.0018* (-1.733) (-1.731) (-1.895) CASH -0.0001 -0.0001 -0.0002 (-0.413) (-0.406) (-0.798) DUAL -0.0006* -0.0006* -0.0007** (-1.887) (-1.903) (-2.078) GDP -0.0022 -0.0022 -0.0030 (-1.093) (-1.120) (-1.313) INDUS -0.0052 -0.0052 -0.0088 (-0.999) (-0.996) (-1.480) DT 0.0000 (0.137) DT_inter 0.0005** (2.194) SKILL 0.0005 (0.819) SKILL_inter 0.0028* (1.798) _cons 0.8589*** 0.8593*** 0.9014*** (33.491) (33.498) (30.294) FE Yes Yes Yes Year Yes Yes Yes N 36449 36449 33650 R 2 0.9778 0.9778 0.9738 6.2 Organisational Capability Organisational capability refers to an enterprise's capacity to integrate internal and external resources and coordinate multi-party collaboration through structural design, process optimisation, and institutional safeguards to achieve strategic objectives. Within the TOE framework, the organisational dimension focuses on the enterprise's endogenous management efficacy. Enterprises with high ESG ratings place greater emphasis on environmental sustainability. Those with robust internal controls have more comprehensive management systems and resource-integration capabilities, enabling them to leverage external incentives from AI pilot zone policies more effectively to advance their green transition. This paper predicts that for enterprises with high ESG ratings and strong internal controls, the AI Pilot Zone policy exerts a stronger catalytic effect on green transition. First, this study uses the Huazheng ESG rating to measure corporate ESG performance. As shown in Column (3) of Table 6.2, the coefficient for the AI Pilot Zone's impact on green transition is 0.0020, significant at the 1% level. The coefficient for ESG_inter is 0.0099, both significant at the 10% level. This indicates that the higher a company's ESG score, the stronger the AI pilot zone policy's promotional effect on its green transition. Secondly, this study employs the DIB internal control index for measurement. As shown in column (3) of Table 6.2, the coefficient for the impact of AI pilot zones on green transition is 0.0023, significant at the 1% level. The coefficient for IC_inter is 0.0051, significant at the 5% level. This indicates that the higher a company's internal control level, the stronger the promotional effect of AI pilot zone policies on its green transition. Table 6.2 Organisational Capability Heterogeneity (1) ESG Internal Control CGT CGT CGT inter 0.0021*** 0.0020*** 0.0023*** (5.107) (4.789) (5.259) ROA 0.0002 0.0002 -0.0024 (0.427) (0.434) (-1.455) SIZE -0.0000 -0.0000 -0.0002 (-0.161) (-0.115) (-0.658) LEV 0.0001 0.0001 -0.0004 (0.314) (0.229) (-0.671) TOP -0.0009 -0.0009 -0.0001 (-0.632) (-0.617) (-0.084) MBOARD -0.0016* -0.0016* -0.0018* (-1.733) (-1.732) (-1.841) CASH -0.0001 -0.0001 0.0001 (-0.413) (-0.364) (0.273) DUAL -0.0006* -0.0006* -0.0008** (-1.887) (-1.912) (-2.379) GDP -0.0022 -0.0021 -0.0030 (-1.093) (-1.071) (-1.410) INDUS -0.0052 -0.0052 -0.0069 (-0.999) (-0.986) (-1.219) ESG -0.0012 (-0.644) ESG_inter 0.0099* (1.918) IC -0.0006 (-0.720) IC_inter 0.0051** (2.195) _cons 0.8589*** 0.8630*** 0.8782*** (33.491) (32.344) (31.321) FE Yes Yes Yes Year Yes Yes Yes N 36449 36449 32558 R 2 0.9778 0.9778 0.9781 6.3 External Environment Within the environmental dimension, the external environment reflects the intensity of external policy constraints and guidance, directly influencing the momentum and directional choices of green transition. The higher the level of regional industry competition, the more inclined enterprises are to drive technological innovation to maintain market competitiveness. Simultaneously, the higher the level of regional environmental regulation, the greater the external environmental pressure enterprises face. This paper predicts that the greater the level of regional industry competition and the higher the level of regional environmental regulation at an enterprise's location, the stronger the AI Pilot Zone's promotional effect on green transition. First, this study uses the number of concluded intellectual property (IP) cases adjudicated by municipal people's courts in the Peking University Legal Database as a proxy for a city's IP adjudication volume (Shen & Huang, 2019). To account for city scale effects, GDP is used for scale adjustment. Concurrently, to compare the intensity of intellectual property protection across cities, this paper constructs a measure of city-level IP protection intensity using the Revealed Comparative Advantage (RCA) index. The level of intellectual property protection is calculated as: (Local IP case dispositions / Local GDP) / (National IP case dispositions / National GDP)(Blazsek & Escribano, 2016). The results, as shown in Column (1) of Table 6.3, indicate that the coefficient for the impact of AI pilot zone policies on green transition is 0.0012, significant at the 1% level. The coefficient for IPP_inter is 0.0012, significant at the 5% level. This demonstrates that a robust intellectual property system enhances the effectiveness of AI pilot zone policies, thereby more effectively empowering enterprises in their green transition. Secondly, environmental regulations drive technological upgrading through policy standards (Yan et al., 2024). This study measures regional environmental regulation intensity using the frequency proportion of environmental terms (e.g., ‘environmental protection,’ ‘pollution,’ ‘emission reduction’) in local government work reports (Chen & Chen, 2018), . Specifically, this is calculated as the proportion of words in sentences containing environmental terms relative to the total word count of the government work report. The results, as shown in Column (2) of Table 6.3, indicate that the coefficient for the impact of AI pilot zone policies on green transition is 0.0033, significant at the 1% level. The coefficient on ER_inter is 1.1249 and is significant at the 5% level. This demonstrates that stronger regional environmental regulation strengthens the promotional effect of AI pilot zone policies on corporate green transition. Table 6.3 External Environment Heterogeneity (1) Intellectual Property Protection Environmental Regulations CGT CGT CGT inter 0.0021*** 0.0012*** 0.0033*** (5.107) (2.652) (4.317) ROA 0.0002 0.0002 0.0004 (0.427) (0.484) (0.764) SIZE -0.0000 0.0000 -0.0001 (-0.161) (0.053) (-0.263) LEV 0.0001 0.0001 0.0001 (0.314) (0.493) (0.547) TOP -0.0009 -0.0011 -0.0002 (-0.632) (-0.768) (-0.127) MBOARD -0.0016* -0.0018** -0.0014 (-1.733) (-2.005) (-1.377) CASH -0.0001 -0.0001 -0.0003 (-0.413) (-0.600) (-0.976) DUAL -0.0006* -0.0005 -0.0005 (-1.887) (-1.611) (-1.324) GDP -0.0022 -0.0019 -0.0025 (-1.093) (-0.961) (-1.196) INDUS -0.0052 -0.0070 -0.0074 (-0.999) (-1.301) (-1.212) IPP -0.0004 (-1.508) IPP_inter 0.0012** (2.374) ER 0.2511** (2.139) ER_inter 1.1249** (2.123) _cons 0.8589*** 0.8601*** 0.8670*** (33.491) (33.186) (30.919) FE Yes Yes Yes Year Yes Yes Yes N 36449 36269 28981 R 2 0.9778 0.9780 0.9772 7. Research Findings and Policy Implications 7.1 Research Findings This study employs a quasi-natural experiment leveraging the AI Pilot Zone policy, utilising data from A-share listed companies in Shanghai and Shenzhen between 2010 and 2023. A double-difference model was constructed to systematically estimate the causal impact of AI development on corporate green transition behaviour. The findings indicate that the construction of AI pilot zones significantly promotes enterprise green transition, a conclusion that remains valid after a series of robustness tests. The mechanism analysis further reveals that AI policies promote enterprises' green transition by enhancing their green investment levels, improving credit availability and green innovation efficiency, and reducing agency costs. Heterogeneity analyses based on the TOE framework show that the above facilitation effects are more pronounced in firms with higher levels of digital transformation, a sufficient pool of high-tech talent, stronger internal governance, and better ESG performance. In addition, the facilitating effect of AI policies on firms' green transition is more pronounced in regions with stronger intellectual property protection systems and more stringent environmental regulations. Overall, this paper provides important empirical evidence from the micro-firm level that AI policies promote the green transition and enriches research on the economic effects of AI policies and firms' environmental behaviour. 7.2 Policy Recommendations Based on the empirical results of this paper, the AI pilot zone significantly promotes green transition of enterprises through multiple mechanisms such as alleviating financing constraints, enhancing green innovation efficiency, and optimising corporate governance structure, and the effects of the policies show significant heterogeneity under the differences in enterprise resource endowment and institutional environment. ccordingly, the following policy implications are proposed. Firstly, the development of AI pilot zones should be continuously advanced, with spatial planning of intelligent infrastructure optimised. Findings indicate that AI infrastructure and technological applications are crucial exogenous drivers of corporate green transition. Consequently, building on existing achievements in pilot zones, efforts should accelerate the adoption of AI technologies in high-carbon sectors such as manufacturing. Mechanisms for cross-regional technology diffusion and policy coordination should be fostered, with institutional trials and demonstration projects reinforcing AI’s long-term guiding role in green transition. Secondly, the financing support system for green transition should be strengthened, with a focus on alleviating capital constraints faced by enterprises for green investment. Empirical evidence indicates that AI policies significantly promote corporate green transition by improving access to credit and reducing financing constraints. Consequently, efforts should be made to establish an integrated 'AI + green finance' development model. Financial institutions should be encouraged to develop credit products linked to green technology adoption and emission reduction performance. Government-guided funds, risk compensation mechanisms, and interest subsidies should be employed to reduce financing costs for green investments, thereby enhancing the efficiency of financial resource allocation towards corporate green transition. Once again, the whole chain support system for green science and technology innovation should be strengthened to enhance green innovation productivity. Research has found that artificial intelligence promotes green transition of enterprises by improving the efficiency of green research and development and the efficiency of results transformation. Therefore, the policy level should focus on supporting enterprises, universities and research institutions to collaborate on key green technology research, reducing innovation input costs through tax incentives and special funds on the R&D side, improving technology trading platforms and industrialisation service systems on the results transformation side, accelerating the diffusion of green technologies and their commercial application, and promoting the formation of a technological innovation-driven green transition model. Furthermore, differentiated policy supply should be implemented, taking full account of differences in enterprise capacity endowment and regional institutional environments. Heterogeneity analysis shows that enterprises with higher digitalisation levels, sufficient human capital reserves, sound corporate governance and better ESG performance are more likely to realise green transition under the impetus of AI policies. Therefore, classification support policies should be implemented for enterprises at different stages of development, and capacity building for enterprises with weak digitalisation foundations should be strengthened while cultivating model enterprises for green transition. At the same time, the intellectual property protection and environmental regulation systems should be improved to fully leverage the synergistic amplification effect between the institutional environment and AI policies. In addition, the incentive and constraint mechanisms for green transition should be strengthened to enhance enterprises' endogenous motivation for green development. Through the establishment of an evaluation system centred on emission reduction performance and green innovation output, financial and tax incentives can be given to enterprises with remarkable green transition results, and more binding emission regulation and technological transformation requirements can be implemented for high-pollution and high-energy-consumption enterprises, so as to promote the formation of long-term green investment expectations by enterprises. Finally, the supporting system for green transition and information disclosure system should be improved to enhance the effectiveness of policy implementation and resource allocation efficiency. It should accelerate the establishment of a unified enterprise carbon emission statistics and information disclosure system, build a digital low-carbon transformation information platform, and establish a dynamic assessment mechanism for green transition performance in the AI pilot zone, forming a closed-loop management system for policy implementation, effect feedback and system optimisation. Declarations Ethical Approval: This article does not contain any studies with human participants performed by any of the authors. Informed Consent: This article does not contain any studies with human participants performed by any of the authors. Data Availability statement This paper uses A-share listed companies from 2010 to 2023 as the research sample, removes missing observations, and shrinks the data by 1%, yielding 36,449 observations. The main explanatory variables, the list of AI pilot test zones, are from the Ministry of Industry and Information Technology (MIIT), the data at the enterprise level are from the WIND and the Cathay Pacific CSMAR database, the data at the city level are from the City Statistical Yearbook, and the data processing and regression analyses are done with Stata 18.0. they can be accessed from supplementary files. Author Contribution Wenrui Ma （First Author）: Conceptualization, Data Curation, Formal Analysis, Investigation, Methodology, Software, Visualization, Writing-Original Draft, Writing-Review & Editing References Acemoglu, D., & Restrepo, P. (2018). Artificial intelligence, automation, and work. In The economics of artificial intelligence: An agenda (pp. 197-236). University of Chicago Press. https://doi.org/10.7208/chicago/9780226613475.003.0008 Adnan, Z. H., Chakraborty, K., Bag, S., & Wu, J. S. (2025). Pricing and green investment strategies for electric vehicle supply chain in a competitive market under different channel leadership. Annals of Operations Research , 355 (1), 87-122. https://doi.org/10.1007/s10479-023-05523-y Agrawal, A., Gans, J. S., & Goldfarb, A. (2019). Exploring the impact of artificial intelligence: Prediction versus judgment. Information Economics and Policy , 47 , 1-6. https://doi.org/10.1016/j.infoecopol.2019.05.001 Agrawal, A., McHale, J., & Oettl, A. (2018). Finding needles in haystacks: Artificial intelligence and recombinant growth. In The economics of artificial intelligence: An agenda (pp. 149-174). University of Chicago Press. https://doi.org/10.7208/chicago/9780226613475.003.0005 Aktas, N., Andreou, P. C., Karasamani, I., & Philip, D. (2019). CEO Duality, Agency Costs, and Internal Capital Allocation Efficiency. British Journal of Management , 30 (2), 473-493. https://doi.org/10.1111/1467-8551.12277 Babina, T., Fedyk, A., He, A., & Hodson, J. (2024). Artificial intelligence, firm growth, and product innovation. Journal of Financial Economics , 151 , 103745. https://doi.org/10.1016/j.jfineco.2023.103745 Barker Iii, V. L., & Duhaime, I. M. (1997). Strategic change in the turnaround process: Theory and empirical evidence. Strategic management journal , 18 (1), 13-38. https://doi.org/10.1002/(SICI)1097-0266(199701)18:13.0.CO;2-X Berlin, M., & Mester, L. J. (1998). On the profitability and cost of relationship lending. Journal of Banking & Finance , 22 (6-8), 873-897. https://doi.org/10.1016/S0378-4266(98)00033-8 Bia, M., Huber, M., & Lafférs, L. (2024). Double Machine Learning for Sample Selection Models. Journal of Business & Economic Statistics , 42 (3), 958-969. https://doi.org/10.1080/07350015.2023.2271071 Blazsek, S., & Escribano, A. (2016). Patent propensity, R&D and market competition: Dynamic spillovers of innovation leaders and followers. Journal of Econometrics , 191 (1), 145-163. https://doi.org/10.1016/j.jeconom.2015.10.005 Cai, X., Lin, P., & Wang, R. (2025). Intelligent manufacturing and corporate green transition. International Review of Financial Analysis , 97 , 103796. https://doi.org/10.1016/j.irfa.2024.103796 Chang, J., & Yongjian, L. (2025). Artificial Intelligence and Natural Resource Dependency: Impact Effects and Moderation Mechanisms. Statistics and Decision Making , 41 (17), 48-52. https://doi.org/10.13546/j.cnki.tjyjc.2025.17.008 Chen, G. J., Ding, S. J., Zhao, X. Q., & Jiang, X. Y. (2021). China’s green finance policy, financing costs and firms’ green transition: A central bank collateral framework perspective. Journal of Financial Research , 12 (12), 75-95. Chen, H., Yu, Z., & Hu, S. (2025). Digital economy, human capital accumulation, and corporate green total factor productivity: Based on strategic emerging industries. International Review of Financial Analysis , 103 , 104152. https://doi.org/10.1016/j.irfa.2025.104152 Chen, K. C. W., Chen, Z., & Wei, K. C. J. (2009). Legal protection of investors, corporate governance, and the cost of equity capital. Journal of Corporate Finance , 15 (3), 273-289. https://doi.org/10.1016/j.jcorpfin.2009.01.001 Chen, S., & Golley, J. (2014). ‘Green’ productivity growth in China's industrial economy. Energy Economics , 44 , 89-98. https://doi.org/10.1016/j.eneco.2014.04.002 Chen, S. Y., & Chen, D. K. (2018). Air pollution, government regulations and high-quality economic development. Econ. Res. J , 53 (02), 20-34. Chen, X., Chen, F., Liu, X., & Zhao, M. (2024). The impact of industrial robot uses on the labor cost stickiness of Chinese firms. Journal of Accounting Literature . https://doi.org/10.1108/JAL-07-2023-0127 Chen, Y., Liu, M., & Su, J. (2013). Greasing the wheels of bank lending: Evidence from private firms in China. Journal of Banking & Finance , 37 (7), 2533-2545. https://doi.org/10.1016/j.jbankfin.2013.02.002 Chen, Y., & Ma, Y. (2021). Does green investment improve energy firm performance? Energy Policy , 153 , 112252. https://doi.org/10.1016/j.enpol.2021.112252 Chen, Z., Li, X., & Xia, X. (2021). Measurement and spatial convergence analysis of China’s agricultural green development index. Environmental Science and Pollution Research , 28 (16), 19694-19709. https://doi.org/10.1007/s11356-020-11953-z Chong, T. T.-L., Lu, L., & Ongena, S. (2013). Does banking competition alleviate or worsen credit constraints faced by small-and medium-sized enterprises? Evidence from China. Journal of Banking & Finance , 37 (9), 3412-3424. https://doi.org/10.1016/j.jbankfin.2013.05.006 De Faria, P., Lima, F., & Santos, R. (2010). Cooperation in innovation activities: The importance of partners. Research Policy , 39 (8), 1082-1092. https://doi.org/10.1016/j.respol.2010.05.003 De Haas, R., Martin, R., Muûls, M., & Schweiger, H. (2025). Managerial and financial barriers to the green transition. Management science , 71 (4), 2890-2921. https://doi.org/10.1287/mnsc.2023.00772 Du, K., & Li, J. (2019). Towards a green world: How do green technology innovations affect total-factor carbon productivity. Energy Policy , 131 , 240-250. https://doi.org/10.1016/j.enpol.2019.04.033 Fan, F., Lian, H., Liu, X., & Wang, X. (2021). Can environmental regulation promote urban green innovation Efficiency? An empirical study based on Chinese cities. Journal of Cleaner Production , 287 , 125060. https://doi.org/10.1016/j.jclepro.2020.125060 Fang, X. (2024). Can climate risk exposure compel companies to undergo a green transition? Journal of Cleaner Production , 434 , 140310. https://doi.org/10.1016/j.jclepro.2023.140310 Flammer, C. (2021). Corporate green bonds. Journal of Financial Economics , 142 (2), 499-516. https://doi.org/10.1016/j.jfineco.2021.01.010 Frank, M. R., Autor, D., Bessen, J. E., Brynjolfsson, E., Cebrian, M., Deming, D. J., Feldman, M., Groh, M., Lobo, J., & Moro, E. (2019). Toward understanding the impact of artificial intelligence on labor. Proceedings of the National Academy of Sciences , 116 (14), 6531-6539. https://doi.org/10.1073/pnas.1900949116 Geng, C., & Wen, G. (2025). Reform of the social credit system, green innovation, and corporate resilience. Finance Research Letters , 108550. https://doi.org/10.1016/j.frl.2025.108550 Guan, T., Zheng, R., & Chen, A. (2026). Artificial intelligence and corporate energy consumption: The policy effects of the new-generation artificial intelligence innovation and development pilot zones. Economic Analysis and Policy , 89 , 148-164. https://doi.org/10.1016/j.eap.2025.11.032 Han, S., Sun, X., & Zhong, S. (2025). Applications of artificial intelligence and urban innovation performance: A quasi-natural experiment based on the pilot zones for the innovative application of artificial intelligence. Plos one , 20 (8), e0329729. https://doi.org/10.1371/journal.pone.0329729 Herrera, A. M., & Minetti, R. (2007). Informed finance and technological change: Evidence from credit relationships. Journal of Financial Economics , 83 (1), 223-269. https://doi.org/10.1016/j.jfineco.2005.12.001 Hilb, M. (2020). Toward artificial governance? The role of artificial intelligence in shaping the future of corporate governance. Journal of Management and Governance , 24 (4), 851-870. https://doi.org/10.1007/s10997-020-09519-9 Huang, L., & Lei, Z. (2021). How environmental regulation affect corporate green investment: Evidence from China. Journal of Cleaner Production , 279 , 123560. https://doi.org/10.1016/j.jclepro.2020.123560 Humberd, B. K., & Latham, S. F. (2025). When AI becomes an agent of the firm: examining the evolution of AI in organizations through an agency theory lens. Journal of Management Studies . https://doi.org/10.1111/joms.13274 Huseynov, F., & Klamm, B. K. (2012). Tax avoidance, tax management and corporate social responsibility. Journal of Corporate Finance , 18 (4), 804-827. https://doi.org/10.1016/j.jcorpfin.2012.06.005 Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. The American economic review , 76 (2), 323-329. Jia, N., Luo, X., Fang, Z., & Liao, C. (2024). When and how artificial intelligence augments employee creativity. Academy of Management Journal , 67 (1), 5-32. https://doi.org/10.5465/amj.2022.0426 Jiakui, C., Abbas, J., Najam, H., Liu, J., & Abbas, J. (2023). Green technological innovation, green finance, and financial development and their role in green total factor productivity: Empirical insights from China. Journal of Cleaner Production , 382 , 135131. https://doi.org/10.1016/j.jclepro.2022.135131 Jiang, T. (2022). Mediating effects and moderating effects in causal inference. China Industrial Economics , 5 (5), 100-120. https://doi.org/10.19581/j.cnki.ciejournal.2022.05.005. Jiang, W., Huang, X., Gan, L., & Martek, I. (2025). Dynamic pricing and green investment in the construction supply chain with incentive and consumer reference price. Engineering, Construction and Architectural Management , 1-29. https://doi.org/10.1108/ecam-04-2024-0489 Kemp, R., & Never, B. (2017). Green transition, industrial policy, and economic development. Oxford Review of Economic Policy , 33 (1), 66-84. https://doi.org/10.1093/oxrep/grw037 Kim, S., & Yoon, A. (2023). Analyzing active fund managers’ commitment to ESG: Evidence from the United Nations Principles for Responsible Investment. Management science , 69 (2), 741-758. https://doi.org/10.1287/mnsc.2022.4394 Kuang, C. E., Liu, J. Y., & Li, W. Y. (2024). Digital-intelligence integration empowerment and green transition of manufacturing enterprises. Contemporary Finance & Economics , 5 , 114-127. Lee, K. Y., Jung, H. J., & Kwon, Y. (2024). Boundary-spanning technology search, product component reuse, and new product innovation: Evidence from the smartphone industry. Research Policy , 53 (4), 104959. https://doi.org/10.1016/j.respol.2024.104959 Li, B., Peng, X., & Ouyang, M. (2013). Environmental regulation, green total factor productivity and the transformation of China’s industrial development mode—Analysis based on data of China’s 36 industries. China Ind. Econ , 4 , 56-68. https://doi.org/10.19581/j.cnki.ciejournal.2013.04.005 Li, D., & Wang, P. (2025). FinTech development and corporate green total factor productivity. International Review of Financial Analysis , 107 , 104567. https://doi.org/10.1016/j.irfa.2025.104567 Li, P., Yang, D., Li, P., Ye, Z., & Deng, Z. (2011). A study on the green transition of Chinese industry. China Industrial Economics , 4 , 5-14. https://doi.org/10.19581/j.cnki.ciejournal.2011.04.001 Li, T., Meng, X., Wang, L., & Gong, Y. (2025). Will China's Carbon Emission Trading Regulation Really Decrease the Green Investment of Firms? A Microperspective on Corporate Governance. Business strategy and the environment , 34 (2), 2222-2238. https://doi.org/10.1002/bse.4103 Li, W., Hu, H., & Hong, Z. (2024). Green finance policy, ESG rating, and cost of debt——Evidence from China. International Review of Financial Analysis , 92 , 103051. https://doi.org/10.1016/j.irfa.2023.103051 Lin, B.-q., & Yang, Y.-x. (2025). Building efficiency: How the national AI innovation pilot zones enhance green energy utilization? Evidence from China. Journal of Environmental Management , 387 , 125945. https://doi.org/10.1016/j.jenvman.2025.125945 Liu, C., Cui, L., & Li, C. (2022). Impact of environmental regulation on the green total factor productivity of dairy farming: Evidence from China. Sustainability , 14 (12), 7274. https://doi.org/10.3390/su14127274 Liu, H., Tan, M., & Cao, R. (2026). The role of AI in promoting urban sustainable development: Evidence from China's new generation of national AI innovative development pilot zones. Applied Geography , 188 , 103906. https://doi.org/10.1016/j.apgeog.2026.103906 Liu, Q., & Zhao, C. (2020). Tax incentives and upgrading firms’ human capital. Economic Research Journal , 55 (4), 70-85. Liu, Y., Wu, K., & Liang, X. (2024). Does low-carbon pilot policy promote corporate green total factor productivity? Economic Analysis and Policy , 84 , 1-24. https://doi.org/10.1016/j.eap.2024.08.015 Maghyereh, A., Boulanouar, Z., & Essid, L. (2025). The dynamics of green innovation and environmental policy stringency in energy transition investments. Journal of Cleaner Production , 487 , 144649. https://doi.org/10.1016/j.jclepro.2024.144649 Mäkelä, H., & Näsi, S. (2010). Social responsibilities of MNCs in downsizing operations: A Finnish forest sector case analysed from the stakeholder, social contract and legitimacy theory point of view. Accounting, Auditing & Accountability Journal , 23 (2), 149-174. Manzaneque, M., Rojo-Ramírez, A. A., Diéguez-Soto, J., & Martínez-Romero, M. J. (2020). How negative aspiration performance gaps affect innovation efficiency. Small Business Economics , 54 (1), 209-233. https://doi.org/10.1007/s11187-018-0091-8 Martin, G. P., Wiseman, R. M., & Gomez-Mejia, L. R. (2019). The interactive effect of monitoring and incentive alignment on agency costs. Journal of Management , 45 (2), 701-727. https://doi.org/10.1177/0149206316678453 Meng, X., Zhang, M., & Zhao, Y. (2024). Environmental regulation and green transition: Quasi-natural experiment from China's efforts in sulfur dioxide emissions control. Journal of Cleaner Production , 434 , 139741. https://doi.org/10.1016/j.jclepro.2023.139741 Mijit, R., Hu, Q., Xu, J., & Ma, G. (2025a). Greening through AI? The impact of Artificial Intelligence Innovation and Development Pilot Zones on green innovation in China. Energy Economics , 108507. https://doi.org/10.1016/j.eneco.2025.108507 Mijit, R., Hu, Q., Xu, J., & Ma, G. (2025b). Greening through AI? The impact of Artificial Intelligence Innovation and Development Pilot Zones on green innovation in China. Energy Economics , 146 , 108507. https://doi.org/10.1016/j.eneco.2025.108507 Oh, D.-h. (2010). A global Malmquist-Luenberger productivity index. Journal of productivity analysis , 34 (3), 183-197. https://doi.org/10.1007/s11123-010-0178-y Pang, S., Hua, G., & Liu, Y. (2025). Legitimacy and Opportunism: Examining Climate Risk and Green Mergers and Acquisitions in Polluting Firms. Business Ethics, the Environment & Responsibility . https://doi.org/10.1111/beer.12829 Qian, S., Yang, Z., Yang, L., & Zhang, Y. (2025). Institutional investors’ green attention and corporate greenwashing: The effectiveness of external governance. Economic Analysis and Policy . https://doi.org/10.1016/j.eap.2025.05.048 Raisch, S., & Fomina, K. (2025). Combining human and artificial intelligence: Hybrid problem-solving in organizations. Academy of Management Review , 50 (2), 441-464. https://doi.org/10.5465/amr.2021.0421 Sacconi, L. (2006). A social contract account for CSR as an extended model of corporate governance (I): Rational bargaining and justification. Journal of business ethics , 68 (3), 259-281. https://doi.org/10.1007/s10551-006-9014-8 Shahzad, M. F., Xu, S., An, X., Asif, M., & Jafri, M. A. H. (2024). Effect of stakeholder pressure on environmental performance: Do virtual CSR, green credit, environmental and social reputation matter? Journal of Environmental Management , 368 , 122223. https://doi.org/10.1016/j.jenvman.2024.122223 Shen, G., & Huang, S. (2019). The impact of city-level intellectual property protection on foreign capital entry into Chinese enterprises. Finance and trade economies , 40 , 143-157. Shen, Y., & Zhang, X. (2024). The impact of artificial intelligence on employment: the role of virtual agglomeration. Humanities and Social Sciences Communications , 11 (1), 122. https://doi.org/10.1057/s41599-024-02647-9 Shi, L., Cai, Z., Ding, X., Di, R., & Xiao, Q. (2020). What Factors Affect the Level of Green Urbanization in the Yellow River Basin in the Context of New-Type Urbanization? Sustainability , 12 (6), 2488. https://doi.org/10.3390/su12062488 Tang, D., & Yang, Z. (2022). Local environmental expenditures, fiscal subsidies on environmental protection and corporate green technology innovation. Public Fin Res , 467 , 79-93. Tian, J., Sun, S., Cao, W., Bu, D., & Xue, R. (2024). Make every dollar count: The impact of green credit regulation on corporate green investment efficiency. Energy Economics , 130 , 107307. https://doi.org/10.1016/j.eneco.2024.107307 Tian, P., & Lin, B. (2017). Promoting green productivity growth for China's industrial exports: evidence from a hybrid input-output model. Energy Policy , 111 , 394-402. https://doi.org/10.1016/j.enpol.2017.09.033 Tone, K. (2001). A slacks-based measure of efficiency in data envelopment analysis. European journal of operational research , 130 (3), 498-509. https://doi.org/10.1016/S0377-2217(99)00407-5 Torrent‐Sellens, J., Enache‐Zegheru, M., & Ficapal‐Cusí, P. (2025). Promoting the European Sustainable Firm: How Economic, Social, and Green Innovation and the AI‐Based Technologies Create Pathways of Social and Environmental Sustainability. Business strategy and the environment , 34 (7), 9093-9119. https://doi.org/10.1002/bse.70068 Vanneste, B. S., & Puranam, P. (2024). Artificial intelligence, trust, and perceptions of agency. Academy of Management Review (ja), amr-2022. Wang, A., Bi, C., & Zhang, Y. (2025). Government accounting supervision and supply chain environmental investment. Finance Research Letters , 76 , 106977. https://doi.org/10.1016/j.frl.2025.106977 Wang, H., Guo, T., & Tang, Q. (2021). The effect of national culture on corporate green proactivity. Journal of Business Research , 131 , 140-150. https://doi.org/10.1016/j.jbusres.2021.03.023 Wang, K., Sun, K., Li, Y., & Xie, Q. (2025). Can digital technology break the financing dilemma of innovative SMEs? Based on the perspective of enterprise innovation. Technological Forecasting and Social Change , 214 , 124030. https://doi.org/10.1016/j.techfore.2025.124030 Wang, S., Wang, Y., & Li, C. (2024). AI-driven capital-skill complementarity: Implications for skill premiums and labor mobility. Finance Research Letters , 68 , 106044. https://doi.org/10.1016/j.frl.2024.106044 Wang, X. Q., & Ning, J. H. (2020). Can mandatory social responsibility disclosure drive corporate green transition? Evidence based on green patent data of listed companies in China. Journal of Audit & Economics , 35 (4), 69-77. Wu, F., Hu, H., Lin, H., & Ren, X. (2021). Enterprise digital transformation and capital market performance: Empirical evidence from stock liquidity. Management world , 37 (7), 130-144. Yan, S., Almandoz, J., & Ferraro, F. (2021). The impact of logic (in) compatibility: Green investing, state policy, and corporate environmental performance. Administrative Science Quarterly , 66 (4), 903-944. https://doi.org/10.1177/00018392211005756 Yan, Z., Yu, Y., Du, K., & Zhang, N. (2024). How does environmental regulation promote green technology innovation? Evidence from China's total emission control policy. Ecological Economics , 219 , 108137. https://doi.org/10.1016/j.ecolecon.2024.108137 Yang, Y., & Chi, Y. (2023). Path selection for enterprises’ green transition: Green innovation and green mergers and acquisitions. Journal of Cleaner Production , 412 , 137397. https://doi.org/10.1016/j.jclepro.2023.137397 Yu, F., & Chen, J. (2023). The impact of industrial internet platform on green innovation: Evidence from a quasi-natural experiment. Journal of Cleaner Production , 414 , 137645. https://doi.org/10.1016/j.jclepro.2023.137645 Yu, X. L., & Zhou, Y. (2023). Green Credit Policy and Green Transition of High-pollution Enterprises: A Perspective from Emission Reduction and Development. The Journal of Quantitative & Technical Economics , 5 , 1-22. https://doi.org/10.19581/j.cnki.ciejournal.2011.04.001 Yuan, B., & Xiang, Q. (2018). Environmental regulation, industrial innovation and green development of Chinese manufacturing: Based on an extended CDM model. Journal of Cleaner Production , 176 , 895-908. https://doi.org/10.1016/j.jclepro.2017.12.034 Yuxin, R., & Zhengchu, H. (2025). Urban Digital Intelligence Development and Corporate New Quality Productive Forces: A Quasi-natural Experiment Based on National Artificial Intelligence (AI) Innovative Development Pilot Zones. Journal of Shanghai University of Finance and Economics , 27 (3), 34-48. https://doi.org/10.16538/j.cnki.jsufe.2025.03.003 Zhai, X., An, Y., Shi, X., & Liu, X. (2022). Measurement of green transition and its driving factors: Evidence from China. Journal of Cleaner Production , 335 , 130292. https://doi.org/10.1016/j.jclepro.2021.130292 Zhang, L., & Zhang, S. (2020). Technology Empowerment: The Effect of Technological Innovation on the Integrated Development of Artificial Intelligence and Industry. Financ. Econ , 15 . Zhang, R. (2024). Fintech-driven green economy leapfrog: an exploration of jiangsu province paths based on entropy weighted TOPSIS-coupled coordination models. Environment, Development and Sustainability , 1-33. https://doi.org/10.1007/s10668-024-05809-6 Zhang, S., Cheng, L., Ren, Y., & Yao, Y. (2024). Effects of carbon emission trading system on corporate green total factor productivity: Does environmental regulation play a role of green blessing? Environmental research , 248 , 118295. https://doi.org/10.1016/j.envres.2024.118295 Zhang, S., Dong, R., Jiang, J., Yang, S., Cifuentes-Faura, J., Peng, S., & Feng, Y. (2024). Whether the green credit policy effectively promote green transition of enterprises in China? Empirical analysis and mechanism verification. Environmental research , 244 , 117910. https://doi.org/10.1016/j.envres.2023.117910 Zhang, W., & Sun, C. (2023). How does outward foreign direct investment affect enterprise green transition? Evidence from China. Journal of Cleaner Production , 428 , 139331. https://doi.org/10.1016/j.jclepro.2023.139331 Zhang, X., Jin, Y., & Shen, C. (2020). Manufacturers’ green investment in a competitive market with a common retailer. Journal of Cleaner Production , 276 , 123164. https://doi.org/10.1016/j.jclepro.2020.123164 Zhao, M., Gao, Y., Feng, J., & Sun, W. (2024). Effect test of China’s outward foreign direct investment reverse technology spillover on green total factor production under cultivation of new quality productivity forces. Bulletin of Chinese Academy of Sciences (Chinese Version) , 40 (2), 350-361. Zhao, X., Nakonieczny, J., Jabeen, F., Shahzad, U., & Jia, W. (2022). Does green innovation induce green total factor productivity? Novel findings from Chinese city level data. Technological Forecasting and Social Change , 185 , 122021. https://doi.org/10.1016/j.techfore.2022.122021 Zhou, Z., Nie, L., Ji, H., Zeng, H., & Chen, X. (2020). Does a firm’s low-carbon awareness promote low-carbon behaviors? Empirical evidence from China. Journal of Cleaner Production , 244 , 118903. https://doi.org/10.1016/j.jclepro.2019.118903 Zhu, L., Li, T., Wang, C., & Huang, J. (2025). Corporate ESG performance and green innovation: moderating effect of shareholder activism. Journal of Environmental Management , 383 , 125413. https://doi.org/10.1016/j.jenvman.2025.125413 Zhu, X., Liu, R., & Chen, J. (2023). Corporate environmental investment and supply chain financing: The moderating role of environmental innovation. Business strategy and the environment , 32 (4), 1559-1581. https://doi.org/10.1002/bse.3205 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 15 Apr, 2026 Reviews received at journal 30 Mar, 2026 Reviews received at journal 26 Mar, 2026 Reviewers agreed at journal 26 Mar, 2026 Reviewers agreed at journal 20 Mar, 2026 Reviewers agreed at journal 18 Mar, 2026 Reviewers agreed at journal 18 Mar, 2026 Reviewers invited by journal 17 Mar, 2026 Editor assigned by journal 17 Mar, 2026 Editor invited by journal 11 Mar, 2026 Submission checks completed at journal 09 Mar, 2026 First submitted to journal 05 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8993678","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":608346804,"identity":"ec4d0217-80a1-4201-af5b-747fbd99b985","order_by":0,"name":"Wenrui Ma","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9UlEQVRIie3Rv0vEMBTA8RcO4lJ0fUVo/4UUoZ7gH5Nw0C4erhkLhXQpzvdndDtu63HgLe8P0C0iOFdc3DQRJ4fU8YZ8p/zgAy8EIBY7wVizOk5S8Ey4pdsvf0+DpKrA6vOrHzICzhOAu5JZytTg1/8ii06WVpmy3l4e2o9JI1x0I083ocF6WwtlqvXuQRkcCQFJ8nQIkY0cUZnH9UDMwN64wZ6ApzZIVOPIVy2ItZMn+TxZcZTEpSMNeiI8CQ7Wvy2E1LzY9cwgESYFqfYm9Pyiu39/+RQ8v07OXietb7PseNg/9yHS/DlIZj8yD97GYrFYzPcNXk1Sdr90b4EAAAAASUVORK5CYII=","orcid":"","institution":"University of Bristol","correspondingAuthor":true,"prefix":"","firstName":"Wenrui","middleName":"","lastName":"Ma","suffix":""}],"badges":[],"createdAt":"2026-02-28 09:09:28","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8993678/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8993678/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105052809,"identity":"7c5de655-06fc-42e3-996c-3cc5f387b758","added_by":"auto","created_at":"2026-03-20 10:48:21","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":86198,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCGT Parallel Trend Test\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8993678/v1/bc692eebf122d0b686846ef1.png"},{"id":105562993,"identity":"f8069941-fb0a-4a7c-b135-746c580d098c","added_by":"auto","created_at":"2026-03-27 12:45:33","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":76091,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCGT Placebo Test\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8993678/v1/23311a03a778fdf5c39cde58.png"},{"id":105569161,"identity":"5a741248-0668-4e6d-9f07-a26f16191636","added_by":"auto","created_at":"2026-03-27 13:11:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2115982,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8993678/v1/bce235b2-efc0-4577-81ec-4ccf88fad85f.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Artificial Intelligence and Corporate Green Transition--Evidence From China's AI Innovation Development Pilot Zones","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eWith the global rise of the green economy, the call for green industrial development has grown increasingly urgent. As key players in industrial sectors and primary contributors to carbon emissions and pollution control, enterprises' green development has become a vital driving force for optimising industrial structures, enhancing productivity levels, and achieving high-quality economic growth.\u003c/p\u003e \u003cp\u003eIn China, all sectors of society have come to recognise the imperative to shift away from the previous extensive development model, characterised by 'high input, high consumption, and high pollution'. This requires driving the green transition of the economy to achieve high-quality development. In 2022, the report to the 20th National Congress of the Communist Party of China emphasized the need to \u0026lsquo;promote the high-end, intelligent, and green upgrading of traditional industries, and build a green, low-carbon, circular economic system\u0026rsquo;. Furthermore, the 2025 \u0026lsquo;Proposal of the Central Committee of the Communist Party of China on Formulating the 15th Five-Year Plan for National Economic and Social Development' proposed 'promoting technological transformation and upgrading, advancing the digital and intelligent transformation of manufacturing, developing smart manufacturing, green manufacturing, and service-oriented manufacturing, and accelerating the transformation of industrial models and enterprise organisational forms.'\u003c/p\u003e \u003cp\u003eGuided by national strategy, China's enterprises have now entered a new phase of systematic advancement and scaled development in green growth. Regarding the green manufacturing system, by 2025, China had cumulatively cultivated 6,430 green factories, with their output accounting for over 20% of total manufacturing output. In green innovation, the State Intellectual Property Office's Green and Low-Carbon Patent Statistical Analysis Report (2025) indicates that China's low-carbon patent inventions totaled 120,000 in 2024, accounting for nearly 50% of global growth. Chinese enterprises dominate green innovation, accounting for over 70% of green patent applications, and are driving vigorous green and low-carbon technological innovation. Furthermore, in corporate sustainability disclosure, 2,532 A-share listed companies published ESG-related reports in 2025, achieving a disclosure rate of approximately 46.58% and a cumulative net increase of 20% over five years. However, constrained by factors such as imperfect corporate management systems, difficulties in penetrating the green technology market, a lack of clear green standards, and obstacles to industrial-chain emission reductions, the current advancement and effectiveness of corporate green transition are hindered. Effectively promoting corporate green transition not only enhances sustainable development levels but is also pivotal to driving national industrial transformation and comprehensive green development across the economy and society.\u003c/p\u003e \u003cp\u003eIn recent years, with the advent of a new wave of technological revolution and the deepening of industrial transformation trends, artificial intelligence has become a vital engine for the deep integration of traditional production factors and data elements (Zhang \u0026amp; Zhang, \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). As an inherent component of modern industrial systems, artificial intelligence is a generalised science encompassing methods, theories, and applied systems designed to simulate and augment human intelligence. The technologies (products) it generates are regarded as pivotal drivers of the \u0026lsquo;Fourth Industrial Revolution\u0026rsquo;. Currently, disruptive technologies centred on AI continue to give rise to new industries, new models and new kinetic energy. China has taken several measures to promote AI applications and technological innovation and has released the New Generation Artificial Intelligence Development Plan in 2017. On 21 May 2019, the Ministry of Industry and Information Technology (MIIT) and the Shanghai Municipal People's Government (Municipal People's Government) jointly inaugurated the first pilot zone in Shanghai, the \"Shanghai ( Pudong New Area) Pilot Zone for Artificial Intelligence Innovation and Application\", marking the official landing of the National Pilot Zone for Artificial Intelligence Innovation and Application Plan (hereinafter collectively referred to as the AI Policy). Concurrently, in October of the same year, Shenzhen and Qingdao joined Shanghai as the first batch of pilot cities. In February 2021, the second cohort of pilot cities included Beijing, Tianjin, Hangzhou, Guangzhou, and Chengdu. By September 2022, Nanjing, Wuhan, and Changsha were approved as pilot cities. As of February 2026, eleven cities nationwide had been authorised to establish pilot zones. Functioning as core platforms for incubating and scaling intelligent solutions, these AI pilot zones foster regional resource coordination and integration through a dual-track approach of policy support and market-driven initiatives. This facilitates the development of technology validation and application scenarios(Mijit et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2025a\u003c/span\u003e). Moreover, within the pilot zone, the introduction of artificial intelligence technology enables enterprises to significantly enhance resource utilisation efficiency, reduce environmental pollution, improve production sustainability, and overcome challenges in the green transition.\u003c/p\u003e \u003cp\u003eTaking the AI pilot zone policy as the research entry point, this paper constructs a quasi-natural experiment to identify the causal impact of AI development on the green transition behaviour of enterprises, which expands the existing research in terms of research perspectives, identification strategies and mechanism analysis. Firstly, this paper systematically examines the impact of AI on enterprise green transition from a policy perspective, breaking through the limitations of existing studies that mainly analyse it from the perspectives of digital technology application or enterprise digital transformation. Second, this paper leverages the exogenous policy shock from the establishment of AI pilot zones and employs a multi-period double-difference model to mitigate potential endogeneity, providing empirical evidence with greater causal explanatory power for identifying the promotion of AI for green transition. Again, this paper constructs a multiple transmission mechanism of financing constraint alleviation, green innovation efficiency enhancement and agency cost reduction, revealing the intrinsic role path of AI in promoting green transition of enterprises. Finally, using the Technology-Organisation-Environment (TOE) analysis framework, this paper systematically examines the heterogeneity of policy effects across the dimensions of enterprise resource endowment and institutional environment, deepening understanding of the interaction mechanism between AI and green transition.\u003c/p\u003e"},{"header":"2. Literature Review","content":"\u003cp\u003eAt a time when global environmental problems are becoming increasingly severe, green transition of enterprises has become a core issue of environmental governance in China and even globally. In this context, how to fully leverage the subversive and revolutionary technological advantages of artificial intelligence to empower enterprises\u0026apos; green transition has become a subject of growing concern in the academic community.\u003c/p\u003e\n\u003cp\u003eIn the green transition of enterprises, the existing literature primarily focuses on three aspects: conceptual definitions, measurement methods, and factors affecting it. In terms of defining the concept of green transition, the Chinese Academy of Social Sciences (CASS) suggests that green transition is a process in which industries move towards \u0026apos;intensive use of energy and resources, reduction of pollutant emissions, reduction of environmental impacts, improvement of labour productivity, and enhancement of the capacity for sustainable development\u0026rsquo; (Li et al., 2011). In this process, the green transition of enterprises is more evident in the shift in production mode from high energy consumption and high emissions to low energy consumption and low emissions (Kemp \u0026amp; Never, 2017). This paper combines the views of other scholars to further deepen the connotation of green transition, which is understood as a development model that requires enterprises to consider both economic and environmental performance, and to achieve ecological improvement and green economic and social development through industrial upgrading as well as technological innovation (Meng et al., 2024).\u003c/p\u003e\n\u003cp\u003eIn measuring enterprises\u0026apos; green transition, existing studies primarily focus on composite index methods, quantitative text analysis methods, single-indicator methods, and DEA methods. For instance, numerous scholars employ the Analytic Hierarchy Process (AHP), linear weighting methods (Z. Chen et al., 2021), entropy weighting(Zhang, 2024), or principal component analysis (PCA) combined with factor analysis (Shi et al., 2020) to construct comprehensive index measurement frameworks. These incorporate multiple factors, such as economic transformation, technological transformation, and the energy transition (Zhai et al., 2022), into assessments of industrial and regional green transitions; Some scholars, adopting a corporate strategic orientation perspective, extract keyword frequencies such as \u0026apos;carbon\u0026apos;, \u0026apos;carbon dioxide emissions\u0026apos;, and \u0026apos;greenhouse gas (GHG) emissions\u0026apos; from corporate annual reports and social responsibility reports. These are then scored on a 0-5 scale to quantify corporate green transition behaviour (Zhou et al., 2020); Additionally, researchers have employed proxy variables such as reductions in corporate carbon emissions (Li et al., 2013), the proportion of revenue from polluting industries relative to the top five revenue streams (Yang \u0026amp; Chi, 2023), and investment in green transition projects (Wang et al., 2021)for direct measurement. The aforementioned approaches neglect the social benefits associated with resource and environmental factors(Chen \u0026amp; Golley, 2014; Tian \u0026amp; Lin, 2017) and exhibit issues such as highly subjective indicator selection and limited measurement dimensions, thereby failing to reflect the substantive level of corporate green transition. Consequently, numerous scholars have built upon traditional DEA models, employing super-efficient SBM and GML models (Oh, 2010; Tone, 2001) to further incorporate energy consumption and environmental pollution-related conceptual indicators (Li et al., 2013) into total factor productivity (TFP) calculations. These approaches also introduce two categories of variables: undesirable outputs and energy inputs(Zhao et al., 2024), thereby constructing a green total factor productivity (GTFP) measurement model. As an indicator better suited to evaluating levels of corporate green transition, green TFP\u0026mdash;through objective weighting, simultaneous handling of multiple inputs and outputs, and efficiency diagnostics\u0026mdash;enables more systematic and precise measurement of relative efficiency and improvement pathways. Consequently, it has garnered increasing academic attention(Du \u0026amp; Li, 2019; Jiakui et al., 2023; Zhao et al., 2022). Numerous studies have employed it as an explanatory variable to investigate the impact of smart manufacturing(Cai et al., 2025), fintech(Li \u0026amp; Wang, 2025), digital transformation, the digital economy (Chen et al., 2025), and environmental policies (Liu et al., 2024; Shaopeng Zhang et al., 2024). To avoid misjudging firms\u0026apos; levels of green transition, this paper uses green total factor productivity as a proxy.\u003c/p\u003e\n\u003cp\u003eRegarding the factors influencing the green transition, most scholars have examined the inhibitory or promotional effects of various elements on corporate green transition from three perspectives: policy influence, environmental regulation, and intelligentisation. Firstly, with the state\u0026apos;s increasing focus on the green transition of development models, a series of policy interventions have been shown to exert significant incentive effects on corporate green transitions. These measures extend beyond direct incentives such as green credit (Yu \u0026amp; Zhou, 2023), environmental R\u0026amp;D subsidies (Tang \u0026amp; Yang, 2022), and tax and fee reductions (Huseynov \u0026amp; Klamm, 2012), but also extend to institutional innovations such as green finance (G. J. Chen et al., 2021), carbon emissions trading (Li et al., 2025), and industrial internet pilot programmes (Yu \u0026amp; Chen, 2023). Secondly, corporate willingness to pursue green transition is often constrained by profit maximisation considerations (Maghyereh et al., 2025). Against this backdrop, environmental regulations emerge as a driving factor for corporate green transition. For instance, mandatory social responsibility disclosures and clean production industry technical standards strengthen oversight of corporate legitimacy motives, impose stricter requirements on pollution emission intensity, and thereby advance corporate green transition(Wang \u0026amp; Ning, 2020). However, in the long term, environmental regulations may crowd out corporate R\u0026amp;D investment, thereby suppressing innovation capabilities and hindering green transition (Yuan \u0026amp; Xiang, 2018). Overall, environmental regulations exhibit positive, negative, and non-linear relationships with corporate green transition (Liu et al., 2022). Thirdly, from an intelligent perspective, digital and intelligent technologies enhance corporate green awareness and incentivise green innovation by integrating digital and intelligent applications, thereby empowering green transition (Kuang et al., 2024). Additionally, scholars examining industrial robot imports note that while such robots boost production capacity, they also enable more sustainable emission-reduction pathways through green manufacturing, thereby advancing the integration of economic and environmental benefits (Torrent‐Sellens et al., 2025). Furthermore, specific climate risks, such as water scarcity and resource depletion, significantly compel companies to intensify their research and development efforts. This drives the development of green, intelligent supply chains, thereby compelling enterprises to undergo green transition (Fang, 2024).\u003c/p\u003e\n\u003cp\u003eThe National Pilot Zone for Artificial Intelligence Innovation and Application is a demonstration zone approved by China\u0026apos;s Ministry of Industry and Information Technology to promote the deep integration of artificial intelligence with the real economy. The construction of the pilot zone is driven by the opening of application scenarios, creating a curator of AI innovation and an industrial highland through core technology research, the layout of arithmetic infrastructure, and the cultivation of an industrial ecosystem. Since 2019, Shanghai, Shenzhen, and Jinan-Qingdao have been designated as the first batch of national AI innovation and application pilot zones. As of 2025, 11 cities across the country have been included in the pilot zones, covering Shanghai, Shenzhen, Jinan-Qingdao, Beijing, Tianjin, Hangzhou, Guangzhou, Chengdu, Nanjing, Wuhan, and Changsha. Based on their respective advantages, the zones focus on intelligent manufacturing, smart cities, healthcare, autonomous driving, and other areas, promoting the deep integration of AI and the real economy. Based on regional resource endowments, the pilot regions have formed a differentiated development pattern, as shown in Table 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1: Overview of China\u0026apos;s National AI Innovation\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eDevelopment\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Pilot Zones\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"602\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBatch\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eApproval Date\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePilot Zone Name\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 319px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDistinctive Application Fields and Development Priorities\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eFirst Batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eMay/October 2019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eShanghai (Pudong)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eFocuses on AI chips (Zhangjiang), smart finance (Lujiazui), intelligent manufacturing and smart healthcare. Promoting deep integration of AI algorithms, chips and application scenarios.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eFirst Batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eMay/October 2019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eShenzhen\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eLeveraging its robust electronics industry, prioritising AI chips, smart hardware (particularly smart terminals), smart finance and AI-assisted urban governance.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eFirst Batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eMay/October 2019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eJinan-Qingdao\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eThis dual-core pilot zone sees Jinan focusing on smart healthcare and intelligent software; Qingdao, leveraging enterprises like Haier, concentrates on industrial internet and smart home systems while developing smart marine technologies.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eSecond batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eFebruary 2021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eBeijing\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eCapitalizing on its top-tier scientific talent, prioritises algorithm innovation and security, smart cities (Urban Brain), smart government services, and intelligent connected vehicles.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eSecond batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eFebruary 2021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eTianjin (Binhai)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eLeveraging its port and manufacturing strengths, it prioritises the development of smart ports (Tianjin Port), intelligent manufacturing, and IT application innovation.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eSecond batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eFebruary 2021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eHangzhou\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eLeveraging its digital economy strengths (e.g., Alibaba), prioritising development of City Brain (origin), smart retail/e-commerce, smart finance (FinTech), and smart healthcare.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eSecond batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eFebruary 2021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eGuangzhou\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eCapitalizing on its commercial and automotive industry foundations, focusing on intelligent connected vehicles, smart logistics, smart cities (transport governance), and smart healthcare.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eSecond batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eFebruary 2021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eChengdu\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eLeveraging its status as a western hub for healthcare and cultural tourism, prioritising smart healthcare (Huaxi Hospital, etc.), smart cultural tourism, smart finance, and smart agriculture.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eThird batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eSeptember 2022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eNanjing\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eCapitalising on its robust software industry foundation, focusing on intelligent software, smart chips, intelligent manufacturing, and smart grids.。\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eThird batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eSeptember 2022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eWuhan\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eLeveraging the \u0026lsquo;China Optics Valley\u0026rsquo; and scientific-educational resources, prioritising development in intelligent manufacturing (particularly intelligent optoelectronics), intelligent remote sensing, intelligent connected vehicles, and smart healthcare.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003eThird batch\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 122px;\"\u003e\n \u003cp\u003eSeptember 2022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 89px;\"\u003e\n \u003cp\u003eChangsha\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 319px;\"\u003e\n \u003cp\u003eCapitalising on its distinctive industries, focusing on intelligent construction machinery (e.g., Sany Heavy Industry), intelligent connected vehicles, and smart cultural/creative industries/media (Malan Mountain).\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eExisting research primarily examines the economic effects of artificial intelligence pilot zones from three perspectives: labour reallocation, corporate innovation, and corporate green development. First, artificial intelligence may reshape skill requirements and occupational opportunities, raising concerns about large-scale technological unemployment(Frank et al., 2019). Against this backdrop, pilot zone development has significantly promoted labour reallocation from services to manufacturing by raising entry barriers for low-skilled workers into high-end productive services while reducing skill requirements in manufacturing, thereby enhancing labour resource allocation efficiency (Wang et al., 2024). The establishment of pilot zones also supports employment at both the urban and enterprise levels through industrial agglomeration, expanded corporate markets, and job creation effects (Shen \u0026amp; Zhang, 2024), thereby promoting high-quality full employment. Second, corporate innovation is often constrained by factors such as uncertain returns (K. Wang et al., 2025), policy instability (Barker Iii \u0026amp; Duhaime, 1997), and the innovation performance expectation gap (Manzaneque et al., 2020). The establishment of pilot zones, through stable policy mechanisms, enhances corporate profitability, financing capacity, and technological capabilities, thereby supporting both overall innovation and breakthrough innovation (Fan et al., 2021). Additionally, scholars note that artificial intelligence pilot zones help enterprises secure greater government digital subsidies, reduce profit uncertainty, and attract regional digital talent and capital, thereby creating favourable conditions for corporate innovation and digital transformation (Han et al., 2025; Yuxin \u0026amp; Zhengchu, 2025). Furthermore, Razia found that the development of pilot zones significantly promotes the adoption of artificial intelligence within enterprises(Mijit et al., 2025b). Enhancing technological integration in turn elvates green innovation performance, collectively fostering sustainable innovation growth. Third, green development is a core tenet of the new development philosophy. From an environmental-constraints perspective, Jin conducted an empirical analysis using Chinese provincial panel data, which indicates that establishing AI pilot zones significantly reduced provincial dependence on natural resources (Chang \u0026amp; Yongjian, 2025). This was achieved by enhancing corporate technological efficiency and adjusting industrial structures, thereby further realising corporate green development. Lin, adopting an energy structure optimisation perspective, further demonstrates that AI pilot zone policies markedly enhance energy utilisation efficiency in pilot cities (Lin \u0026amp; Yang, 2025). This effect is particularly pronounced where urban environmental regulation, economic development, and infrastructure construction are already at leading levels, ultimately advancing the realisation of corporate green development objectives.\u003c/p\u003e\n\u003cp\u003eExisting studies have mainly focused on the impact of policy environment, environmental regulation and digital technology on the green transition of enterprises; in the field of artificial intelligence, studies have mainly focused on its role in the productivity, technological innovation and governance structure of enterprises, but there is a lack of systematic analyses of the causal impacts and multiple mechanisms of artificial intelligence in promoting the green transition of enterprises. In addition, the synergistic role of enterprises\u0026apos; internal resource endowment and external institutional environment in green transition has not been fully examined. To address the above shortcomings, this paper takes the AI pilot zone policy as a quasi-natural experiment, systematically identifies the causal impact of AI on enterprise green transition, and analyses the role paths from the multi-dimensional perspective of financing constraints, green innovation efficiency, and governance mechanisms, and at the same time examines the heterogeneity between enterprises and the institutional environment based on the TOE framework, so as to enrich the existing research in the level of policy identification and mechanism integration.\u003c/p\u003e"},{"header":"3. Theoretical Analysis and Research Hypotheses","content":"\u003cp\u003eBased on stakeholder theory, signaling theory, social learning theory and principal-agent theory, this paper argues that the AI Innovation and Application Pilot Zone policy can promote green transition of enterprises by enhancing the level of green investment, credit availability, green innovation efficiency as well as reducing the agency costs.\u003c/p\u003e \u003cp\u003eFirstly, from the perspective of corporate green investment: on the one hand, grounded in stakeholder theory, artificial intelligence as a novel general-purpose technology can optimise production processes and enhance productivity, thereby reducing costs, increasing profits, and satisfying management's short-term performance incentives, thus providing endogenous funding for green investment. Simultaneously, AI continuously generates high-frequency, multidimensional operational and energy-consumption data. Green investors, as stakeholders, can utilise this data to identify a firm's green transition potential, thereby increasing exogenous green investment support. Furthermore, AI can directly empower green investment decision-making by predicting energy demand and evaluating project returns alongside environmental benefits, thereby reducing investment uncertainty and strengthening management's willingness to invest. On the other hand, increased green investment will significantly drive corporate green transition. From a financing perspective, green investment directly funds corporate green transition projects, alleviating financing constraints (Zhang \u0026amp; Sun, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). From a governance perspective, green investment often entails specific environmental performance targets or market preferences (Yan et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), guiding enterprises to optimise product portfolios by increasing R\u0026amp;D investment and production share of green products (Chen \u0026amp; Ma, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and establish green supply chains (Adnan et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). These actions elevate corporate green transition levels on the supply side, meeting market expectations. In addition, based on the perspective of external supervision, as the scale of green investment expands, external investors such as green funds and banks will be involved in corporate management decisions (Kim \u0026amp; Yoon, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), which not only strengthens the supervision of corporate behaviours, but also significantly inhibits \u0026lsquo;greenwashing\u0026rsquo; (Qian et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and opportunistic tendencies (Pang et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), ensuring that funds really flow to green transition activities, encouraging shareholder activism (Zhu et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)to oversee firms to strengthen green supply chain management and improve green innovation to realise the green transition (A. Wang et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Zhu et al., \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSecond, from the perspective of enterprise credit availability, on the one hand, in traditional financial practices, enterprises mainly rely on establishing close ties with banks to obtain credit (Berlin \u0026amp; Mester, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1998\u003c/span\u003e), and enterprise credit availability is subject to the double constraints of information asymmetry between banks and enterprises and the transaction costs of loans (Chong et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), which results in the green transition activities of enterprises being hindered by the problems of 'difficult financing\u0026rsquo; and \u0026lsquo;expensive financing\u0026rsquo;. To address the above problems, based on signaling theory, AI technology can help banks to effectively capture enterprise internal operation information by promoting the intelligence and dataisation of enterprise production process, so that enterprises can leave enough \u0026lsquo;digital footprints\u0026rsquo;, such as payment records and logistic information (Babina et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and thus help banks to effectively capture enterprise internal operation information (Babina et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), which can help banks to effectively capture enterprise internal operation information (Agrawal et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Frank et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Hilb, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and reduce the information asymmetry between banks and enterprises; at the same time, AI can transform non-standardised data already available to enterprises into visual information (Agrawal et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Babina et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), simplifying the bank review process, compressing the transaction cost and approval time (Chen et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), reducing the transaction cost of enterprise loans, and thus enhancing credit availability.\u003c/p\u003e \u003cp\u003eOn the other hand, increased credit availability substantially contributes to the green transition. Firstly, improved financing conditions and extended debt repayment cycles provide enterprises with long-term, stable funding for energy-saving upgrades and green innovation (Herrera \u0026amp; Minetti, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Shujing Zhang et al., \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Secondly, sustained and enhanced credit accessibility continuously incentivises enterprises to optimise internal operations and environmental disclosure, deepening green practices (Flammer, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Finally, grounded in legitimacy theory, an implicit social contract exists between enterprises and society (M\u0026auml;kel\u0026auml; \u0026amp; N\u0026auml;si, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Sacconi, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Enhanced credit accessibility attracts greater stakeholder scrutiny, generating contractual pressure (De Haas et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). To safeguard their reputation and reduce capital costs, firms have incentives to advance green transition in alignment with societal value norms, thereby enhancing legitimacy(Chen et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Shahzad et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFurthermore, from the perspective of corporate green innovation efficiency, artificial intelligence facilitates green transition by elevating the efficacy of green innovation. On the one hand, grounded in social learning theory, corporate green innovation relies on inventors' capacity to absorb external information and engage in mutual intellectual stimulation (De Faria et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Leveraging technological advantages such as machine learning, intelligent perception, and big data analytics, artificial intelligence can proactively process and deeply mine data assets to generate innovative knowledge insights (Jia et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Consequently, inventors can establish novel social networks characterised by heightened innovative thinking through human-machine collaboration (Babina et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Leveraging the broader knowledge spillover effects generated by such networks (Raisch \u0026amp; Fomina, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), this approach enhances cross-firm and cross-industry information exchange, thereby facilitating technological convergence (Mijit et al., \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2025b\u003c/span\u003e), significantly enhancing corporate efficiency in green innovation product development and technology commercialisation rates (Lee et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Moreover, artificial intelligence can elevate corporate demand for highly skilled labour through \"machine substitution for human labour\" (Acemoglu \u0026amp; Restrepo, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), thereby reducing barriers to entry and trial-and-error costs in green innovation, achieving an overall efficiency leap. Conversely, enhanced green innovation efficiency continuously elevates corporate green transition levels. First, grounded in efficiency theory, the synergistic increase in both the quantity and quality of corporate green innovations cultivates technological advantages in green production. This enhances resource utilisation efficiency at the supply end, reduces pollution during production processes, and ultimately guides the green transition of corporate production workflows. Secondly, on the demand side, green product innovation responds to market trends towards low-carbon consumption. Through differentiation strategies, it enhances brand value and reputation (Yang \u0026amp; Chi, \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), thereby generating new business opportunities and deepening firms' willingness and capacity for green transition.\u003c/p\u003e \u003cp\u003eFinally, from the perspective of corporate agency costs: firstly, based on principal-agent theory, these costs arise from the separation of ownership and management rights within firms, comprising principal monitoring costs, agent collateral costs, and residual losses (Jensen, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e1986\u003c/span\u003e). Artificial intelligence technology can mitigate these agency costs by enhancing information transparency through sampling analysis and real-time alerts of vast internal operational data and external public sentiment. Furthermore, AI is entering a new phase of continuous learning and self-improvement, with its behavioural patterns shifting from simple human mimicry toward autonomous decision-making capabilities (Vanneste \u0026amp; Puranam, \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Aligning AI's autonomous decision-making objectives with long-term corporate value not only leverages AI's \"rational agent\" advantage (Humberd \u0026amp; Latham, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), reducing cost stickiness caused by agent self-interest and short-termism (Chen et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), optimising corporate resource allocation. Moreover, more sophisticated predictive models can minimise residual losses stemming from decision biases within principal-agent relationships (Humberd \u0026amp; Latham, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Conversely, reduced agency costs further empower corporate green transitions. Firstly, improved incentive structures align management interests with long-term corporate development. To build sustainable competitive advantages, management demonstrates greater willingness to advance green transition(Li et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Martin et al., \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Secondly, firms will persistently rely on high information transparency to maintain low agency costs, making their environmental performance more susceptible to public scrutiny (Geng \u0026amp; Wen, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Consequently, green transition gradually becomes a voluntary choice for management. Finally, reduced agency costs signify enhanced resource utilisation efficiency (Aktas et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), directing greater resources towards corporate green governance activities and directly empowering green transition.\u003c/p\u003e \u003cp\u003eIn summary, this paper proposes the following hypothesis:\u003c/p\u003e \u003cp\u003eH1: Artificial intelligence effectively drives corporate green transition.\u003c/p\u003e \u003cp\u003eH2: Artificial intelligence facilitates corporate green transition by enhancing green investment levels, credit availability, and the efficiency of green innovation, while reducing agency costs.\u003c/p\u003e"},{"header":"4. Research Design","content":"\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e4.1 Model Specification\u003c/h2\u003e\n \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e\n \u003ch2\u003e4.1.1 Baseline Regression Model\u003c/h2\u003e\n \u003cp\u003eTo examine the impact of artificial intelligence pilot zones on corporate green transition, the following regression model is constructed:\u003c/p\u003e\n \u003cdiv id=\"Equa\" class=\"Equation\"\u003e\n \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\u003cimg src=\"https://myfiles.space/user_files/58895_8739fc6c57c1c19a/58895_custom_files/img1774003189.png\" width=\"755\" height=\"363\"\u003e\u003c/div\u003e\n \u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e\n \u003ch2\u003e4.1.2 Mediating Effect Model\u003c/h2\u003e\n \u003cp\u003eTo explore the mechanism of AI pilot zones\u0026apos; role in enterprises\u0026apos; green transition, this paper adopts a mediation effect analysis. Given that the traditional stepwise regression method for estimating mediated effects suffers from endogeneity bias (Jiang, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), this paper examines the impact of core explanatory variables on mediating variables by replacing the explanatory variables with the mediating variables to test the existence and validity of the mediating path. The mediating effect model is set as follows:\u003c/p\u003e\n \u003cp\u003e\u003cimg src=\"https://myfiles.space/user_files/58895_8739fc6c57c1c19a/58895_custom_files/img1774003086.png\" width=\"755\" height=\"71\"\u003e\u003c/p\u003e\n \u003cp\u003ewhere the mediating variables are credit availability (Flex2), corporate green investment (CGI), green R\u0026amp;D efficiency (GRD), green outcome conversion efficiency (GCON), and agency costs (AC). All other variables are consistent with Eq.\u0026nbsp;(1).\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003e4.2 Variable Selection\u003c/h2\u003e\n \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e\n \u003ch2\u003e4.2.1 Dependent Variables\u003c/h2\u003e\n \u003cp\u003eCorporate Green transition (CGT): Incorporating environmental pollution into the evaluation framework, the non-radial SBM-ML index (hereafter \u0026lsquo;ML index\u0026rsquo;) is employed to measure corporate green total factor productivity (Chen et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). This productivity comprises input and output indicators, with inputs consisting of labour, capital, and energy inputs, while outputs encompass desired and undesired outputs. This paper employs the following methodology to measure input and output indicators: (1) Input factors: Labour input is proxied by the number of enterprise employees; capital input is proxied by net fixed assets; energy input is proxied by the enterprise\u0026apos;s share of industrial electricity consumption in its city\u0026apos;s total urban industrial electricity consumption, calculated based on the proportion of enterprise employees relative to the city\u0026apos;s total urban employed population. (2) Desired output: enterprise business income as a proxy variable for the desired output of the enterprise. (3) Undesired output: the proportion of employees in the enterprise in the employment of urban workers in the city where it is located is converted to the \u0026lsquo;three industrial wastes\u0026apos;, namely, industrial Sulphur dioxide, industrial wastewater and industrial smoke and dust emissions, as a proxy for the enterprise\u0026apos;s undesired output.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e\n \u003ch2\u003e4.2.2 Explanatory Variables\u003c/h2\u003e\n \u003cp\u003eArtificial Intelligence Pilot Zone Policy (inter): as an explanatory variable, the Artificial Intelligence Pilot Zone serves as a core platform for incubating and promoting intelligent solutions and builds AI technology validation and application scenarios through the two-way linkage of policy support and market-driven forces, accumulating replicable and promotable experience for the whole country. In terms of green development, the establishment of AI pilot zones is committed to promoting the in-depth integration of AI technology with green innovation and sustainable development, achieving significant improvement in energy efficiency and optimal allocation of resources through technological innovation, and thus promoting the transformation and development of a low-carbon economy (Guan et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2026\u003c/span\u003e). inter is equal to Treat \u0026times; Post, in which the number of cities that implement the AI The enterprises matched by the cities implementing the pilot test zone policy are assigned a value of 1 as the experimental group; the remaining enterprises are assigned a value of 0 as the control group. The time dummy variable Post is assigned a value of 1 for the first year of implementing the AI pilot zone policy and beyond, and 0 for the rest of the years (Liu et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2026\u003c/span\u003e).\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\n \u003ch2\u003e4.2.3 Mediating Variable\u003c/h2\u003e\n \u003cp\u003eCredit Availability (Flex2): Credit availability is a core indicator of an enterprise\u0026apos;s ability to obtain bank credit resources, and this paper follows the established research paradigm, using the ratio of interest-bearing liabilities to total assets to measure it (Flex2\u0026thinsp;=\u0026thinsp;interest-bearing liabilities/total assets). Among them, the accounting calibre of interest-bearing liabilities is defined as the sum of short-term borrowings, interest payable, non-current liabilities due within one year, long-term borrowings and bonds payable. The higher this indicator\u0026apos;s value, the stronger the enterprise\u0026apos;s ability to obtain credit. As a key link between the financial market and business operations, the level of credit availability has a profound impact on the efficiency of the enterprise\u0026apos;s capital turnover and the potential for investment and development. Especially during the monetary policy adjustment cycle, the high sensitivity of credit availability to changes in the financial environment makes it a key indicator of shifts in corporate financing constraints.\u003c/p\u003e\n \u003cp\u003eCorporate Green Investment (CGI): Green investment can be measured by the amount of investment in environmental protection, i.e., the amount of investment in environmental protection projects in progress (Tian et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) (e.g., desulphurisation and denitrification, wastewater treatment, exhaust gas treatment, clean innovation, procurement of environmental protection equipment, etc.). The measurement of this variable is primarily based on studies of CSR or green investment. By aggregating the various expenditures related to pollution prevention, ecological management and green production in the financial statements of enterprises to derive the total annual green investment, and then dividing it by the total assets at the end of the period for standardisation to get the green investment, it can more accurately measure the strength and scale of the enterprises\u0026apos; investment in green transition. In this paper, we refer to Jiang (Jiang et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and Huang (Huang \u0026amp; Lei, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), and measure the intensity of green investment by the ratio of environmental protection construction in progress/end-of-period total assets of enterprises.\u003c/p\u003e\n \u003cp\u003eGreen science and technology R\u0026amp;D efficiency (GRD) and green achievement conversion efficiency (GCON): drawing on the measurement idea of Bi et al. (2016) (Bi et al., 2016), in the stage of green science and technology R\u0026amp;D, the number of enterprise R\u0026amp;D personnel and R\u0026amp;D expenditure are selected as the initial inputs, and the number of enterprise green patent applications and the number of enterprise green patent authorisations are taken as the intermediate output indicators. In the green results application stage, the intermediate outputs are the number of enterprise green patent applications and the number of enterprise green patent authorisations, while the sales revenue, enterprise pollution emission index and enterprise energy consumption index are used as the final output indicators. The efficiencies of both stages are calculated using the advanced DEA-SBM model.\u003c/p\u003e\n \u003cp\u003eAgency cost (AC): corporate agency cost is a kind of expense due to the separation of corporate ownership and operation, the higher the agency cost, the more serious the agency conflict between management and shareholders. In this paper, the following indicators are selected as proxy variables for corporate agency costs: (1) operating expense ratio (AC1); (2) management expense ratio (AC2); (3) asset turnover ratio (AC3).\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\n \u003ch2\u003e4.2.4 Control Variables\u003c/h2\u003e\n \u003cp\u003eWith reference to previous studies, the following variables that may affect the green transition of enterprises are selected: enterprise size (SIZE): expressed as the logarithmic value of the enterprise\u0026apos;s total assets at the end of the year; shareholding ratio of the first largest shareholder (TOP): expressed as the shareholding share of the first largest shareholder divided by the company\u0026apos;s total share capital; gearing ratio (LEV): expressed as the enterprise\u0026apos;s total liabilities at the end of the year divided by the total assets at the end of the year; profitability of total assets (ROA). Net profit divided by total assets; CASH: Net cash flow from operating activities divided by total assets; DUAL: 1 if the chairman of the board of directors is the same person as the managing director, 0 otherwise; MBOARD, measured by the number of board members. Gross regional product (GDP): Ln(city GDP). Regional industrial upgrading (INDUS) is defined as the value added of the regional tertiary industry divided by the value added of the secondary industry.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003e4.3 Data source and processing\u003c/h2\u003e\n \u003cp\u003eThis paper uses A-share listed companies from 2010 to 2023 as the research sample, removes missing observations, and shrinks the data by 1%, yielding 36,449 observations. The main explanatory variables, the list of AI pilot test zones, are from the Ministry of Industry and Information Technology (MIIT), the data at the enterprise level are from the WIND and the Cathay Pacific CSMAR database, the data at the city level are from the City Statistical Yearbook, and the data processing and regression analyses are done with Stata 18.0. The descriptive statistics of the variables are detailed in Table \u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e4.1\u003c/span\u003e.\u003c/p\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003cbr\u003e\u003c/div\u003e\u0026nbsp;\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 4.1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eDescriptive Statistics\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003ccolgroup cols=\"7\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eVarName\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eObs\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003eSD\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c5\"\u003e\n \u003cp\u003eMin\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c6\"\u003e\n \u003cp\u003eMedian\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c7\"\u003e\n \u003cp\u003eMax\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e1.0281\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.1079\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0.8000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e1.0419\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e1.2553\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.1418\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.3488\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e1.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.0403\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.2381\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e-7.2851\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0.0375\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e29.6619\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e22.2593\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e1.3420\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e13.0760\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e22.0704\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e28.6969\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.4396\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.4366\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0.0080\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0.4205\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e31.4667\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.3386\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.1499\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0.0029\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0.3150\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e0.8999\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e2.1269\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.1975\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0.6931\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e2.1972\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e2.9444\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.2306\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.5209\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e-25.0434\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0.1452\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e13.3110\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.2741\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.4461\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e1.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e10.5984\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.7960\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e6.2400\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e10.6807\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e11.8180\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e2.4792\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.1474\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e2.1323\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e2.4502\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\n \u003cp\u003e2.8461\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003c/div\u003e"},{"header":"5. Empirical Findings","content":"\u003cp\u003e5.1\u0026nbsp;Baseline Regression\u0026nbsp;Results\u003c/p\u003e\n\u003cp\u003eTable 5.1 reports the results of the baseline regression of the AI pilot zone policy on firms\u0026apos; green transition, where column (1) is the regression model without control variables, and columns (2) to (5) are the regression setups with no fixed year and individual, fixed effects of year only, fixed effects of individual only, and fixed effects of year and individual in both directions, respectively. The results show that the regression results in columns (1) to (5) are negative and significant, and the regression coefficient of the core explanatory variable AI pilot test area (inter) in column (5) is 0.0021 and is significant at the 1% level. This indicates that all the AI pilot zone policies significantly promote green transition of enterprises.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5.1 Baseline Regression Results\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0877***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0020***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0257***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(5.183)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(61.398)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(6.278)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(48.000)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(5.107)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-0.0124***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-0.0010*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-6.380)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(0.844)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-1.667)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(0.427)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0144***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0031***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(39.293)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(1.392)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(10.576)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-0.161)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-0.0078***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0013***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-7.190)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(1.036)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(3.586)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(0.314)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-0.0883***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0015**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-0.0298***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-28.481)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-2.352)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-13.792)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-0.632)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-0.0612***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0012**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-0.0074***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-25.290)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-2.471)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-5.681)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-1.733)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0063***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(7.127)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(0.281)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(0.392)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-0.413)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0031***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-0.0022***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(2.970)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-1.098)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-4.700)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-1.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0446***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.2769***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0022\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(74.537)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(0.118)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(276.869)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-1.093)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.0796***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0006\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.2401***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(23.587)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(0.862)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(42.123)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(-0.999)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.8204***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.1880***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.8196***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e-2.5491***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.8589***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(1764.204)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(14.723)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(303.194)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e(-283.374)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e(33.491)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.3630\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.9741\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.9533\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 94px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eNote: ***, ** and * denote significance levels of 1%, 5% and 10% respectively. The same applies to the table below.\u003c/p\u003e\n\u003cp\u003e5.2 Parallel Trends Test\u003c/p\u003e\n\u003cp\u003eThe validity of the double difference model relies on the fulfilment of the parallel trend assumption, i.e. the outcome variables of the experimental group and the control group should show similar trends before the policy intervention, otherwise it may lead to biased estimation of the policy effect. In order to verify this assumption, this paper takes the year before the policy implementation as the base period and constructs a multi-period double difference model to test the trend consistency of the green transition of enterprises, and the results are shown in Figure 1. As can be seen from the figure, at each observation point before the policy intervention (pre_5 to pre_2), the coefficients of enterprise green transition of the experimental group and the control group fluctuate around the value of 0 in a small way, and the corresponding confidence intervals contain 0, which indicates that there is no significant difference in the trend of change of the green transition of the enterprises in the two groups, and that the hypothesis of parallel trend is valid. And after the implementation of the policy (pos_1 and after), the enterprise green transition coefficient changes significantly, indicating that the policy has a significant promotion effect on enterprise green transition.\u003c/p\u003e\n\u003cp\u003e5.3 Placebo Test\u003c/p\u003e\n\u003cp\u003eTo control for potential omitted variables, this paper tests whether the policy effect is due to random factors using a placebo test. Specifically, a randomised treatment group is constructed using a multi-temporal double-difference method: a sample of the same size as the original treatment group is randomly selected, and a treatment group variable and a policy time-point variable are randomly generated. 500 random policy shocks are simulated for the whole sample, and the resulting output, shown in Figure 2, indicates that the coefficients of the random interaction term are centered around 0 and generally smaller than the true estimate of 0.0021. This suggests that the policy effect is significantly weakened after randomisation, supporting the robustness of this paper\u0026apos;s conclusions.\u003c/p\u003e\n\u003cp\u003e5.4 Propensity Score Matching and Robustness Testing\u003c/p\u003e\n\u003cp\u003e5.4.1 Propensity Score Matching Method\u003c/p\u003e\n\u003cp\u003eTo mitigate the interference that sample selection bias may bring to the results of the benchmark regression, this paper adopts the Propensity Score Matching (PSM) method to conduct a robustness test to further verify whether the impact of the AI pilot zone policy on the green transition of enterprises is robust. Specifically, a matched sample is constructed using firms\u0026apos; basic characteristics. Using a 1:3 nearest-neighbour matching approach, firms in the pilot zone are re-paired with those excluded. Regression estimation is then conducted on the matched sample. Table 5.2 reports the regression results after propensity score matching. Columns (1) and (2) show regression coefficients of 0.0026 and 0.0028, respectively, for the policy variable inter on corporate green transition, both significant at the 1% level. In summary, after controlling sample selection bias, the AI pilot zones still exert a significant promotional effect on corporate green transition, further validating the robustness of the baseline regression results.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5.2 Propensity Score Matching Results\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.0026***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.0028***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(3.090)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(3.227)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e-0.0024\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(-0.876)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.0005\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(0.723)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.0017\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(0.745)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e-0.0092**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(-2.148)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e-0.0061***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(-2.672)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(-0.013)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(-0.046)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e-0.0070\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(-0.447)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.0280\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(1.224)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.8205***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.8293***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(293.645)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e(4.581)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e10436\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e10436\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.9662\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e0.9663\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e5.4.2 Controlling for the Impact of the Pandemic\u003c/p\u003e\n\u003cp\u003eTo further validate the robustness of the baseline regression results, this study treats the COVID-19 pandemic as an exogenous shock that may affect corporate green transition and policy implementation outcomes. The 2020 sample year was therefore excluded to control for the pandemic\u0026apos;s interference. The regression results in Column (1) of Table 5.3 show that the regression coefficient for the core explanatory variable is -0.0262, significant at the 5% level. It shows that the role of the AI pilot zone in facilitating companies\u0026apos; green transition remains even after the year of the outbreak is ruled out.\u003c/p\u003e\n\u003cp\u003e5.4.3 Adding control variables\u003c/p\u003e\n\u003cp\u003eTo further control for potential disturbance factors arising from corporate governance structure and business status, this paper introduces the firm\u0026apos;s age (AGE) and growth (GROWTH) as additional control variables in the regression. According to the regression results in Table 5.3, paragraph (2), the regression coefficient for the core explanatory variable INTER is -0.0360 and is significant at the 1% level. It shows that the result of the role of AI pilot zone on green transition of enterprises remains robust.\u003c/p\u003e\n\u003cp\u003e5.4.4 Excluding the effect of non-manufacturing industries\u003c/p\u003e\n\u003cp\u003eIn order to further test the robustness of the baseline regression results, this paper selects manufacturing enterprises as the research sample to carry out the retest. The regression results in column (3) of Table 5.3 show that the estimated coefficient of the core explanatory variables is 0.0016 and is significant at the 1 per cent statistical level. This indicates that even after excluding the sample of non-manufacturing enterprises, establishing AI pilot zones still has a significant promotional effect on enterprises\u0026apos; green transition, and the baseline regression conclusions remain robust.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5.3 Robustness Testing\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eExclude 2020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eAdd control variables\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eExclude non-manufacturing\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0016***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(4.828)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(5.115)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(3.022)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0005\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(0.553)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(0.463)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(0.716)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.151)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.082)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.634)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(0.380)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(0.363)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(0.465)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0011\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0011\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.0017\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.682)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.729)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(0.868)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0019\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.418)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.751)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.580)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0003\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.381)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.464)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.051)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0007**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0009**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-2.102)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.924)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-2.224)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0028\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0024\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.372)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.084)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.938)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0058\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0140**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.074)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.988)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-2.063)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eGROWTH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-1.557)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eAGE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e-0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(-0.415)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.8664***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.8586***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.8850***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(32.451)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(33.474)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e(26.507)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e33287\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e23664\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.9789\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.9767\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e5.4.5 Robustness test using dual machine learning\u003c/p\u003e\n\u003cp\u003eDual machine learning can effectively handle high-dimensional covariates and mitigate endogeneity\u0026nbsp;(Bia et al., 2024), and more accurately estimate the causal impact of the AI pilot zone policy on enterprises\u0026apos; green transition. In this paper, we use a dual machine learning model and apply common algorithms, such as LASSO, random forest, neural network, and gradient boosting, to robustly test the green transition-promoting effect of the AI pilot zone policy. The regression results in columns (1)-(6) of Table 5.4 show that, regardless of the algorithms, the regression coefficients of the green transition of the enterprises by AI pilot zone policy are all significantly positive, further verifying the robustness and reliability of the benchmark regression conclusions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5.4 Comparative results of machine learning methods\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"662\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 99px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 99px;\"\u003e\n \u003cp\u003eLASSO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eRandom Forest\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eNeural network\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eGradient Boosting\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eElastic Net\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eSVM\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 99px;\"\u003e\n \u003cp\u003e0.002\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e0.003\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e0.003\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e0.005\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e0.002\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e0.056\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 99px;\"\u003e\n \u003cp\u003e(0.000)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(0.001)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(0.002)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(0.001)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(0.000)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e(0.001)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003eControl\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 99px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 99px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 99px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eYES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 99px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e5.5 Mediating Effect Analysis\u003c/p\u003e\n\u003cp\u003e5.5.1 Mediating Role of Green Capital Investment (CGI)\u003c/p\u003e\n\u003cp\u003eColumn (2) of Table 5.4 indicates that the coefficient for the impact of AI pilot zones on green capital investment is 1.5083, significant at the 10% level. This suggests AI pilot zones can enhance corporate green capital investment. Green capital investment systematically drives corporate green transition through financing, governance, and oversight. On the one hand, green investment directly funds corporate green projects, alleviating financing constraints (Zhang \u0026amp; Sun, 2023). On the other hand, by leveraging specific environmental performance requirements (Yan et al., 2021), it guides enterprises to optimise product portfolios and supply chains, thereby enhancing green supply capabilities(Adnan et al., 2025; Chen \u0026amp; Ma, 2021; Zhang et al., 2020). simultaneously, green investment encourages external investors to participate in corporate governance, curbing greenwashing and similar practices while overseeing enhanced green innovation to ensure transformational efficacy (Kim \u0026amp; Yoon, 2023; Pang et al., 2025; Qian et al., 2025). Therefore, green investment plays an important role in the transmission between AI pilot zones and enterprises\u0026apos; green transition, supporting the existence of mediating effects.\u003c/p\u003e\n\u003cp\u003e5.5.2 Mediating Role of Credit Accessibility (Flex2)\u003c/p\u003e\n\u003cp\u003eColumn (3) of Table 5.4 shows that the coefficient for the impact of AI pilot zones on green investment is 0.0035, which is significant at the 5% level, indicating that AI pilot zones can improve enterprise credit availability. The improvement of enterprise credit availability can provide a stable long-term financial guarantee for enterprise\u0026apos;s energy-saving renovation and green innovation (Herrera \u0026amp; Minetti, 2007; Shujing Zhang et al., 2024), and motivate enterprises to optimise their internal operation and improve environmental information disclosure, thus deepening their green practices(Flammer, 2021). Thus, credit availability plays an important role in the transmission between AI pilot zones and corporate green transition, validating the existence of mediating effects.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5.5 Mediating Factors of Green Capital Investment and Credit Accessibility\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eFlex2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e1.5083*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0035**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(5.107)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1.840)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(2.219)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e3.3801***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0912***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.427)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(3.355)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-13.725)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-3.6547***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0233***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.161)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-10.017)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(25.377)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e12.3043***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.5238***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.314)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(20.852)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(137.650)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-15.9960***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0209***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.632)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-6.231)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(3.555)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-2.3762\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0057\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.733)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.544)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.619)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.2023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0159***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.413)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.501)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-12.034)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e2.1558***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0003\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(3.904)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.213)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-2.4557\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0078\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.093)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.747)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.020)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e2.0366\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0541***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.999)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.225)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-2.692)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.8589***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e102.6268**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.3170***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(33.491)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(2.366)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-3.206)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e29382\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e31581\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.1051\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.8582\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e5.5.3 Green Innovation Efficiency Mediators\u003c/p\u003e\n\u003cp\u003eIn columns (2) and (3) of Table 5.5, the coefficient of influence of AI pilot zones on the efficiency of green technology R\u0026amp;D is 0.0062, which is significant at the 10% significant level, and the coefficient of influence on the green achievement conversion efficiency is 0.0074, which is significant at the 5% significant level, which suggests that the AI pilot zones can improve the efficiency of green technology R\u0026amp;D and green achievement conversion. According to the efficiency theory, the synergistic improvement of the quantity and quality of green innovation of enterprises can cultivate the advantages of green production technology, promote the greening of the production process by improving resource efficiency and reducing pollution on the supply side, and respond to the trend of low-carbon consumption to enhance brand value and expand market opportunities with differentiated strategies on the demand side, thus systematically deepening the motivation and ability of green transition of enterprises\u0026nbsp;(Yang \u0026amp; Chi, 2023). Therefore, the efficiency of green science and technology R\u0026amp;D and green achievement conversion play important roles in the transmission between the AI pilot zone and the green transition of enterprises, supporting the existence of a mediating effect.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5.6 Mediating Factors of Green Technology R\u0026amp;D and Conversion Efficiency\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eGRD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eGCON\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0062*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0074**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(5.107)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1.754)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(2.137)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0017\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0036\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.427)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.512)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.082)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0015\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0007\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.161)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.852)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.381)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0016\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0006\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.314)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.775)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.269)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0047\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0103\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.632)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.366)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.809)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.733)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.123)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.117)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0010\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0003\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.413)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.529)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.141)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.322)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.003)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0038\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0114\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.093)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.226)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.682)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0261\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0215\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.999)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.584)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.483)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.8589***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.2443\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.2737\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(33.491)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1.123)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1.262)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e33714\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e33826\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.6276\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.6244\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e5.5.4 Mediating Role of Agency Costs (AC)\u003c/p\u003e\n\u003cp\u003eIn columns (2) (3) (4) of Table 5.6, the coefficients of the impact of AI pilot zones on the impact on agency costs are -0.0315 (significant at the 5% level of significance), -0.0237 (significant at the 10% level of significance), and 0.0249 (significant at the 1% level of significance), which suggests that AI pilot zones reduce the agency costs of firms. The reduction of enterprise agency cost empowers green transition through the dual path of governance and resources. First, the highly transparent governance mechanism brought about by the reduction of agency costs can strengthen external scrutiny and supervision of environmental performance and promote the internalisation of green transition within corporate management consciousness\u0026nbsp;(Geng \u0026amp; Wen, 2025). Second, the decrease in agency costs implies that enterprise resource utilisation becomes more efficient, thereby releasing more resources for green governance activities and directly supporting the transformation process\u0026nbsp;(Aktas et al., 2019). Therefore, agency costs play an important role in the transmission between AI pilot zones and corporate green transition, verifying the existence of mediating effects.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5.7 Mediating Factor of Agency Costs\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eAC1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eAC2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eAC3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0315**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0237*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0249***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(5.107)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-2.511)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-1.927)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(3.307)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0151\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0662***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(0.427)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-1.240)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.191)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(9.481)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0685***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0634***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0624***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.161)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-11.735)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-11.043)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-18.639)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0483***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0494***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0175**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(0.314)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(4.038)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(4.195)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(2.562)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0381\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0335\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0206\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.632)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.935)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.840)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0124\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0115\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0327**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-1.733)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(0.502)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(0.474)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(2.324)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.1628***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.1551***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0193***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.413)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-24.870)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-24.270)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(5.225)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0280***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0246***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0095*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-1.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(3.139)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(2.792)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-1.867)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0080\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0014\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.0436\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-1.093)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.432)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.079)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(1.484)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.3183***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.2213**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e-0.1346*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-0.999)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(3.047)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(2.164)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(-1.679)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.8589***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.9921***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.9736***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e1.8498***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(33.491)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(6.014)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(6.014)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e(4.805)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e31824\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e32148\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e32148\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.1732\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.1264\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0.7998\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e"},{"header":"6 Heterogeneity Analysis","content":"\u003cp\u003eBased on the Technology-Organisation-Environment (TOE) analytical framework, this part empirically examines the impact of AI pilot zones on the green transition of enterprises and focuses on the moderating role of the three types of factors: technology, organisation and environment.\u003c/p\u003e\n\u003cp\u003e6.1 Technological Capability\u003c/p\u003e\n\u003cp\u003eTechnological capability denotes the capacity of corporate management and employees to utilise diverse technological resources. Within the TOE framework, the technological dimension focuses on firms\u0026apos; endogenous technological capabilities.\u003c/p\u003e\n\u003cp\u003eLeveraging mature data infrastructure and employees\u0026apos; efficient technological learning abilities, policy resources can more effectively drive optimisation of green production processes and technological innovation practices. This study predicts that for enterprises with high levels of digital transformation and a high proportion of highly skilled personnel, the policy intensity of AI pilot zones has a stronger promotional effect on green transition.\u003c/p\u003e\n\u003cp\u003eFirst, this study defines enterprise digital transformation indicators as core metrics to characterise the synergistic interaction between technical capability and pilot zone policies\u0026nbsp;(Wu et al., 2021). Heterogeneity in digital transformation, as shown in Column (2) of Table 6.1, reveals: the coefficient for the impact of AI pilot zone policies on green transition is 0.0017, significant at the 1% level. The coefficient for DT_inter is 0.0005, significant at the 5% level, indicating that for enterprises with higher levels of digital transformation, the AI pilot zone policy has a stronger promotional effect on green transition.\u003c/p\u003e\n\u003cp\u003eSecondly, drawing upon the methodology of Liu\u0026nbsp;(Liu \u0026amp; Zhao, 2020), this study measures the heterogeneity of the proportion of highly skilled personnel within enterprises using the ratio of highly skilled personnel to total employees. The heterogeneity of the proportion of highly skilled personnel, as shown in column (3) of Table 6.1, reveals that the coefficient for the AI pilot zone policy on green transition is 0.0019, significant at the 1% level. The coefficient for SKILL_inter is 0.0001, significant at the 5% level. This indicates that enterprises with higher levels of digital transformation experience a stronger promotional effect of the AI pilot zone policy on green transition. The results in column (3) of Table 6.1 show that the coefficient for the impact of AI pilot zone policies on green transition is 0.0019, significant at the 1% level. The coefficient for SKILL_inter is 0.0028, significant at the 10% level. This indicates that the greater the proportion of highly skilled personnel within an enterprise, the stronger the promotional effect of AI pilot zone policies on green transition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 6.1 Technological Capability Heterogeneity\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003edigital transformation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003ehighly skilled personnel\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0017***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0019***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(5.107)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(3.758)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(4.364)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0005\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.427)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.437)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.626)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.161)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.176)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.272)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.314)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.302)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.418)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0010\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.632)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.536)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.610)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0018*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.733)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.731)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.895)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.413)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.406)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.798)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0007**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.903)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-2.078)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0030\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.093)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.120)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.313)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0088\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.999)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.996)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.480)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eDT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.137)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eDT_inter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0005**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(2.194)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eSKILL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0005\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.819)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eSKILL_inter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0028*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1.798)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.8589***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.8593***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9014***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(33.491)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(33.498)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(30.294)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e33650\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9738\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e6.2 Organisational Capability\u003c/p\u003e\n\u003cp\u003eOrganisational capability refers to an enterprise\u0026apos;s capacity to integrate internal and external resources and coordinate multi-party collaboration through structural design, process optimisation, and institutional safeguards to achieve strategic objectives. Within the TOE framework, the organisational dimension focuses on the enterprise\u0026apos;s endogenous management efficacy.\u003c/p\u003e\n\u003cp\u003eEnterprises with high ESG ratings place greater emphasis on environmental sustainability. Those with robust internal controls have more comprehensive management systems and resource-integration capabilities, enabling them to leverage external incentives from AI pilot zone policies more effectively to advance their green transition.\u003c/p\u003e\n\u003cp\u003eThis paper predicts that for enterprises with high ESG ratings and strong internal controls, the AI Pilot Zone policy exerts a stronger catalytic effect on green transition.\u003c/p\u003e\n\u003cp\u003eFirst, this study uses the Huazheng ESG rating to measure corporate ESG performance. As shown in Column (3) of Table 6.2, the coefficient for the AI Pilot Zone\u0026apos;s impact on green transition is 0.0020, significant at the 1% level. The coefficient for ESG_inter is 0.0099, both significant at the 10% level. This indicates that the higher a company\u0026apos;s ESG score, the stronger the AI pilot zone policy\u0026apos;s promotional effect on its green transition.\u003c/p\u003e\n\u003cp\u003eSecondly, this study employs the DIB internal control index for measurement. As shown in column (3) of Table 6.2, the coefficient for the impact of AI pilot zones on green transition is 0.0023, significant at the 1% level. The coefficient for IC_inter is 0.0051, significant at the 5% level. This indicates that the higher a company\u0026apos;s internal control level, the stronger the promotional effect of AI pilot zone policies on its green transition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 6.2 Organisational Capability Heterogeneity\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eESG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eInternal Control\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0020***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0023***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(5.107)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(4.789)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(5.259)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0024\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.427)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.434)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.455)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.161)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.115)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.658)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0004\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.314)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.229)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.671)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.632)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.617)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.084)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0018*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.733)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.732)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.841)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.413)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.364)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(0.273)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0008**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.912)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-2.379)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0030\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.093)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.071)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.410)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0069\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.999)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.986)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-1.219)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eESG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.644)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eESG_inter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0099*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(1.918)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eIC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e-0.0006\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(-0.720)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eIC_inter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.0051**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(2.195)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.8589***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.8630***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.8782***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(33.491)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(32.344)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e(31.321)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e32558\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 150px;\"\u003e\n \u003cp\u003e0.9781\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e6.3 External Environment\u003c/p\u003e\n\u003cp\u003eWithin the environmental dimension, the external environment reflects the intensity of external policy constraints and guidance, directly influencing the momentum and directional choices of green transition. The higher the level of regional industry competition, the more inclined enterprises are to drive technological innovation to maintain market competitiveness. Simultaneously, the higher the level of regional environmental regulation, the greater the external environmental pressure enterprises face. This paper predicts that the greater the level of regional industry competition and the higher the level of regional environmental regulation at an enterprise\u0026apos;s location, the stronger the AI Pilot Zone\u0026apos;s promotional effect on green transition.\u003c/p\u003e\n\u003cp\u003eFirst, this study uses the number of concluded intellectual property (IP) cases adjudicated by municipal people\u0026apos;s courts in the Peking University Legal Database as a proxy for a city\u0026apos;s IP adjudication volume\u0026nbsp;(Shen \u0026amp; Huang, 2019). To account for city scale effects, GDP is used for scale adjustment. Concurrently, to compare the intensity of intellectual property protection across cities, this paper constructs a measure of city-level IP protection intensity using the Revealed Comparative Advantage (RCA) index. The level of intellectual property protection is calculated as: (Local IP case dispositions / Local GDP) / (National IP case dispositions / National GDP)(Blazsek \u0026amp; Escribano, 2016). The results, as shown in Column (1) of Table 6.3, indicate that the coefficient for the impact of AI pilot zone policies on green transition is 0.0012, significant at the 1% level. The coefficient for IPP_inter is 0.0012, significant at the 5% level. This demonstrates that a robust intellectual property system enhances the effectiveness of AI pilot zone policies, thereby more effectively empowering enterprises in their green transition.\u003c/p\u003e\n\u003cp\u003eSecondly, environmental regulations drive technological upgrading through policy standards\u0026nbsp;(Yan et al., 2024). This study measures regional environmental regulation intensity using the frequency proportion of environmental terms (e.g., \u0026lsquo;environmental protection,\u0026rsquo; \u0026lsquo;pollution,\u0026rsquo; \u0026lsquo;emission reduction\u0026rsquo;) in local government work reports\u0026nbsp;(Chen \u0026amp; Chen, 2018), . Specifically, this is calculated as the proportion of words in sentences containing environmental terms relative to the total word count of the government work report. The results, as shown in Column (2) of Table 6.3, indicate that the coefficient for the impact of AI pilot zone policies on green transition is 0.0033, significant at the 1% level. The coefficient on ER_inter is 1.1249 and is significant at the 5% level. This demonstrates that stronger regional environmental regulation strengthens the promotional effect of AI pilot zone policies on corporate green transition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 6.3 External Environment Heterogeneity\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eIntellectual Property Protection\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eEnvironmental Regulations\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eCGT\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003einter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e0.0021***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e0.0012***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e0.0033***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(5.107)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(2.652)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(4.317)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eROA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e0.0004\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(0.427)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(0.484)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(0.764)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eSIZE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e-0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e0.0000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(-0.161)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(0.053)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(-0.263)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eLEV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(0.314)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(0.493)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(0.547)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eTOP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e-0.0009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e-0.0011\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e-0.0002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(-0.632)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(-0.768)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(-0.127)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eMBOARD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e-0.0016*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e-0.0018**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e-0.0014\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(-1.733)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(-2.005)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(-1.377)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eCASH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e-0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e-0.0003\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(-0.413)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(-0.600)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(-0.976)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eDUAL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e-0.0006*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e-0.0005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e-0.0005\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(-1.887)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(-1.611)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(-1.324)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eGDP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e-0.0022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e-0.0019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e-0.0025\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(-1.093)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(-0.961)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(-1.196)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eINDUS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e-0.0052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e-0.0070\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e-0.0074\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(-0.999)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(-1.301)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(-1.212)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eIPP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e-0.0004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(-1.508)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eIPP_inter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e0.0012**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(2.374)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eER\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e0.2511**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(2.139)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eER_inter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e1.1249**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(2.123)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e_cons\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e0.8589***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e0.8601***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e0.8670***\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e(33.491)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e(33.186)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e(30.919)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e36449\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e36269\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e28981\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 141px;\"\u003e\n \u003cp\u003e0.9778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e0.9780\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 144px;\"\u003e\n \u003cp\u003e0.9772\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e"},{"header":"7. Research Findings and Policy Implications","content":"\u003cdiv id=\"Sec34\" class=\"Section2\"\u003e \u003ch2\u003e7.1 Research Findings\u003c/h2\u003e \u003cp\u003eThis study employs a quasi-natural experiment leveraging the AI Pilot Zone policy, utilising data from A-share listed companies in Shanghai and Shenzhen between 2010 and 2023. A double-difference model was constructed to systematically estimate the causal impact of AI development on corporate green transition behaviour. The findings indicate that the construction of AI pilot zones significantly promotes enterprise green transition, a conclusion that remains valid after a series of robustness tests. The mechanism analysis further reveals that AI policies promote enterprises' green transition by enhancing their green investment levels, improving credit availability and green innovation efficiency, and reducing agency costs. Heterogeneity analyses based on the TOE framework show that the above facilitation effects are more pronounced in firms with higher levels of digital transformation, a sufficient pool of high-tech talent, stronger internal governance, and better ESG performance. In addition, the facilitating effect of AI policies on firms' green transition is more pronounced in regions with stronger intellectual property protection systems and more stringent environmental regulations. Overall, this paper provides important empirical evidence from the micro-firm level that AI policies promote the green transition and enriches research on the economic effects of AI policies and firms' environmental behaviour.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec35\" class=\"Section2\"\u003e \u003ch2\u003e7.2 Policy Recommendations\u003c/h2\u003e \u003cp\u003eBased on the empirical results of this paper, the AI pilot zone significantly promotes green transition of enterprises through multiple mechanisms such as alleviating financing constraints, enhancing green innovation efficiency, and optimising corporate governance structure, and the effects of the policies show significant heterogeneity under the differences in enterprise resource endowment and institutional environment. ccordingly, the following policy implications are proposed.\u003c/p\u003e \u003cp\u003eFirstly, the development of AI pilot zones should be continuously advanced, with spatial planning of intelligent infrastructure optimised. Findings indicate that AI infrastructure and technological applications are crucial exogenous drivers of corporate green transition. Consequently, building on existing achievements in pilot zones, efforts should accelerate the adoption of AI technologies in high-carbon sectors such as manufacturing. Mechanisms for cross-regional technology diffusion and policy coordination should be fostered, with institutional trials and demonstration projects reinforcing AI\u0026rsquo;s long-term guiding role in green transition.\u003c/p\u003e \u003cp\u003eSecondly, the financing support system for green transition should be strengthened, with a focus on alleviating capital constraints faced by enterprises for green investment. Empirical evidence indicates that AI policies significantly promote corporate green transition by improving access to credit and reducing financing constraints. Consequently, efforts should be made to establish an integrated 'AI\u0026thinsp;+\u0026thinsp;green finance' development model. Financial institutions should be encouraged to develop credit products linked to green technology adoption and emission reduction performance. Government-guided funds, risk compensation mechanisms, and interest subsidies should be employed to reduce financing costs for green investments, thereby enhancing the efficiency of financial resource allocation towards corporate green transition.\u003c/p\u003e \u003cp\u003eOnce again, the whole chain support system for green science and technology innovation should be strengthened to enhance green innovation productivity. Research has found that artificial intelligence promotes green transition of enterprises by improving the efficiency of green research and development and the efficiency of results transformation. Therefore, the policy level should focus on supporting enterprises, universities and research institutions to collaborate on key green technology research, reducing innovation input costs through tax incentives and special funds on the R\u0026amp;D side, improving technology trading platforms and industrialisation service systems on the results transformation side, accelerating the diffusion of green technologies and their commercial application, and promoting the formation of a technological innovation-driven green transition model.\u003c/p\u003e \u003cp\u003eFurthermore, differentiated policy supply should be implemented, taking full account of differences in enterprise capacity endowment and regional institutional environments. Heterogeneity analysis shows that enterprises with higher digitalisation levels, sufficient human capital reserves, sound corporate governance and better ESG performance are more likely to realise green transition under the impetus of AI policies. Therefore, classification support policies should be implemented for enterprises at different stages of development, and capacity building for enterprises with weak digitalisation foundations should be strengthened while cultivating model enterprises for green transition. At the same time, the intellectual property protection and environmental regulation systems should be improved to fully leverage the synergistic amplification effect between the institutional environment and AI policies.\u003c/p\u003e \u003cp\u003eIn addition, the incentive and constraint mechanisms for green transition should be strengthened to enhance enterprises' endogenous motivation for green development. Through the establishment of an evaluation system centred on emission reduction performance and green innovation output, financial and tax incentives can be given to enterprises with remarkable green transition results, and more binding emission regulation and technological transformation requirements can be implemented for high-pollution and high-energy-consumption enterprises, so as to promote the formation of long-term green investment expectations by enterprises.\u003c/p\u003e \u003cp\u003eFinally, the supporting system for green transition and information disclosure system should be improved to enhance the effectiveness of policy implementation and resource allocation efficiency. It should accelerate the establishment of a unified enterprise carbon emission statistics and information disclosure system, build a digital low-carbon transformation information platform, and establish a dynamic assessment mechanism for green transition performance in the AI pilot zone, forming a closed-loop management system for policy implementation, effect feedback and system optimisation.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthical Approval:\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis article does not contain any studies with human participants performed by any of the authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed Consent:\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis article does not contain any studies with human participants performed by any of the authors.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis paper uses A-share listed companies from 2010 to 2023 as the research sample, removes missing observations, and shrinks the data by 1%, yielding 36,449 observations. The main explanatory variables, the list of AI pilot test zones, are from the Ministry of Industry and Information Technology (MIIT), the data at the enterprise level are from the WIND and the Cathay Pacific CSMAR database, the data at the city level are from the City Statistical Yearbook, and the data processing and regression analyses are done with Stata 18.0. they can be accessed from supplementary files.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eWenrui Ma （First Author）: Conceptualization, Data Curation, Formal Analysis, Investigation, Methodology, Software, Visualization, Writing-Original Draft, Writing-Review \u0026amp; Editing\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAcemoglu, D., \u0026amp; Restrepo, P. (2018). Artificial intelligence, automation, and work. In \u003cem\u003eThe economics of artificial intelligence: An agenda\u003c/em\u003e (pp. 197-236). University of Chicago Press. https://doi.org/10.7208/chicago/9780226613475.003.0008 \u003c/li\u003e\n\u003cli\u003eAdnan, Z. H., Chakraborty, K., Bag, S., \u0026amp; Wu, J. S. (2025). Pricing and green investment strategies for electric vehicle supply chain in a competitive market under different channel leadership. \u003cem\u003eAnnals of Operations Research\u003c/em\u003e,\u003cem\u003e 355\u003c/em\u003e(1), 87-122. https://doi.org/10.1007/s10479-023-05523-y \u003c/li\u003e\n\u003cli\u003eAgrawal, A., Gans, J. S., \u0026amp; Goldfarb, A. (2019). Exploring the impact of artificial intelligence: Prediction versus judgment. \u003cem\u003eInformation Economics and Policy\u003c/em\u003e,\u003cem\u003e 47\u003c/em\u003e, 1-6. https://doi.org/10.1016/j.infoecopol.2019.05.001 \u003c/li\u003e\n\u003cli\u003eAgrawal, A., McHale, J., \u0026amp; Oettl, A. (2018). Finding needles in haystacks: Artificial intelligence and recombinant growth. In \u003cem\u003eThe economics of artificial intelligence: An agenda\u003c/em\u003e (pp. 149-174). University of Chicago Press. https://doi.org/10.7208/chicago/9780226613475.003.0005 \u003c/li\u003e\n\u003cli\u003eAktas, N., Andreou, P. C., Karasamani, I., \u0026amp; Philip, D. (2019). CEO Duality, Agency Costs, and Internal Capital Allocation Efficiency. \u003cem\u003eBritish Journal of Management\u003c/em\u003e,\u003cem\u003e 30\u003c/em\u003e(2), 473-493. https://doi.org/10.1111/1467-8551.12277 \u003c/li\u003e\n\u003cli\u003eBabina, T., Fedyk, A., He, A., \u0026amp; Hodson, J. (2024). Artificial intelligence, firm growth, and product innovation. \u003cem\u003eJournal of Financial Economics\u003c/em\u003e,\u003cem\u003e 151\u003c/em\u003e, 103745. https://doi.org/10.1016/j.jfineco.2023.103745 \u003c/li\u003e\n\u003cli\u003eBarker Iii, V. L., \u0026amp; Duhaime, I. M. (1997). Strategic change in the turnaround process: Theory and empirical evidence. \u003cem\u003eStrategic management journal\u003c/em\u003e,\u003cem\u003e 18\u003c/em\u003e(1), 13-38. https://doi.org/10.1002/(SICI)1097-0266(199701)18:1\u0026lt;13::AID-SMJ843\u0026gt;3.0.CO;2-X \u003c/li\u003e\n\u003cli\u003eBerlin, M., \u0026amp; Mester, L. J. (1998). On the profitability and cost of relationship lending. \u003cem\u003eJournal of Banking \u0026amp; Finance\u003c/em\u003e,\u003cem\u003e 22\u003c/em\u003e(6-8), 873-897. https://doi.org/10.1016/S0378-4266(98)00033-8 \u003c/li\u003e\n\u003cli\u003eBia, M., Huber, M., \u0026amp; Laff\u0026eacute;rs, L. (2024). Double Machine Learning for Sample Selection Models. \u003cem\u003eJournal of Business \u0026amp; Economic Statistics\u003c/em\u003e,\u003cem\u003e 42\u003c/em\u003e(3), 958-969. https://doi.org/10.1080/07350015.2023.2271071 \u003c/li\u003e\n\u003cli\u003eBlazsek, S., \u0026amp; Escribano, A. (2016). Patent propensity, R\u0026amp;D and market competition: Dynamic spillovers of innovation leaders and followers. \u003cem\u003eJournal of Econometrics\u003c/em\u003e,\u003cem\u003e 191\u003c/em\u003e(1), 145-163. https://doi.org/10.1016/j.jeconom.2015.10.005 \u003c/li\u003e\n\u003cli\u003eCai, X., Lin, P., \u0026amp; Wang, R. (2025). Intelligent manufacturing and corporate green transition. \u003cem\u003eInternational Review of Financial Analysis\u003c/em\u003e,\u003cem\u003e 97\u003c/em\u003e, 103796. https://doi.org/10.1016/j.irfa.2024.103796 \u003c/li\u003e\n\u003cli\u003eChang, J., \u0026amp; Yongjian, L. (2025). Artificial Intelligence and Natural Resource Dependency: Impact Effects and Moderation Mechanisms. \u003cem\u003eStatistics and Decision Making\u003c/em\u003e,\u003cem\u003e 41\u003c/em\u003e(17), 48-52. https://doi.org/10.13546/j.cnki.tjyjc.2025.17.008 \u003c/li\u003e\n\u003cli\u003eChen, G. J., Ding, S. J., Zhao, X. Q., \u0026amp; Jiang, X. Y. (2021). China\u0026rsquo;s green finance policy, financing costs and firms\u0026rsquo; green transition: A central bank collateral framework perspective. \u003cem\u003eJournal of Financial Research\u003c/em\u003e,\u003cem\u003e 12\u003c/em\u003e(12), 75-95. \u003c/li\u003e\n\u003cli\u003eChen, H., Yu, Z., \u0026amp; Hu, S. (2025). Digital economy, human capital accumulation, and corporate green total factor productivity: Based on strategic emerging industries. \u003cem\u003eInternational Review of Financial Analysis\u003c/em\u003e,\u003cem\u003e 103\u003c/em\u003e, 104152. https://doi.org/10.1016/j.irfa.2025.104152 \u003c/li\u003e\n\u003cli\u003eChen, K. C. W., Chen, Z., \u0026amp; Wei, K. C. J. (2009). Legal protection of investors, corporate governance, and the cost of equity capital. \u003cem\u003eJournal of Corporate Finance\u003c/em\u003e,\u003cem\u003e 15\u003c/em\u003e(3), 273-289. https://doi.org/10.1016/j.jcorpfin.2009.01.001 \u003c/li\u003e\n\u003cli\u003eChen, S., \u0026amp; Golley, J. (2014). \u0026lsquo;Green\u0026rsquo; productivity growth in China\u0026apos;s industrial economy. \u003cem\u003eEnergy Economics\u003c/em\u003e,\u003cem\u003e 44\u003c/em\u003e, 89-98. https://doi.org/10.1016/j.eneco.2014.04.002 \u003c/li\u003e\n\u003cli\u003eChen, S. Y., \u0026amp; Chen, D. K. (2018). Air pollution, government regulations and high-quality economic development. \u003cem\u003eEcon. Res. J\u003c/em\u003e,\u003cem\u003e 53\u003c/em\u003e(02), 20-34. \u003c/li\u003e\n\u003cli\u003eChen, X., Chen, F., Liu, X., \u0026amp; Zhao, M. (2024). The impact of industrial robot uses on the labor cost stickiness of Chinese firms. \u003cem\u003eJournal of Accounting Literature\u003c/em\u003e. https://doi.org/10.1108/JAL-07-2023-0127 \u003c/li\u003e\n\u003cli\u003eChen, Y., Liu, M., \u0026amp; Su, J. (2013). Greasing the wheels of bank lending: Evidence from private firms in China. \u003cem\u003eJournal of Banking \u0026amp; Finance\u003c/em\u003e,\u003cem\u003e 37\u003c/em\u003e(7), 2533-2545. https://doi.org/10.1016/j.jbankfin.2013.02.002 \u003c/li\u003e\n\u003cli\u003eChen, Y., \u0026amp; Ma, Y. (2021). Does green investment improve energy firm performance? \u003cem\u003eEnergy Policy\u003c/em\u003e,\u003cem\u003e 153\u003c/em\u003e, 112252. https://doi.org/10.1016/j.enpol.2021.112252 \u003c/li\u003e\n\u003cli\u003eChen, Z., Li, X., \u0026amp; Xia, X. (2021). Measurement and spatial convergence analysis of China\u0026rsquo;s agricultural green development index. \u003cem\u003eEnvironmental Science and Pollution Research\u003c/em\u003e,\u003cem\u003e 28\u003c/em\u003e(16), 19694-19709. https://doi.org/10.1007/s11356-020-11953-z \u003c/li\u003e\n\u003cli\u003eChong, T. T.-L., Lu, L., \u0026amp; Ongena, S. (2013). Does banking competition alleviate or worsen credit constraints faced by small-and medium-sized enterprises? Evidence from China. \u003cem\u003eJournal of Banking \u0026amp; Finance\u003c/em\u003e,\u003cem\u003e 37\u003c/em\u003e(9), 3412-3424. https://doi.org/10.1016/j.jbankfin.2013.05.006 \u003c/li\u003e\n\u003cli\u003eDe Faria, P., Lima, F., \u0026amp; Santos, R. (2010). Cooperation in innovation activities: The importance of partners. \u003cem\u003eResearch Policy\u003c/em\u003e,\u003cem\u003e 39\u003c/em\u003e(8), 1082-1092. https://doi.org/10.1016/j.respol.2010.05.003 \u003c/li\u003e\n\u003cli\u003eDe Haas, R., Martin, R., Mu\u0026ucirc;ls, M., \u0026amp; Schweiger, H. (2025). Managerial and financial barriers to the green transition. \u003cem\u003eManagement science\u003c/em\u003e,\u003cem\u003e 71\u003c/em\u003e(4), 2890-2921. https://doi.org/10.1287/mnsc.2023.00772 \u003c/li\u003e\n\u003cli\u003eDu, K., \u0026amp; Li, J. (2019). Towards a green world: How do green technology innovations affect total-factor carbon productivity. \u003cem\u003eEnergy Policy\u003c/em\u003e,\u003cem\u003e 131\u003c/em\u003e, 240-250. https://doi.org/10.1016/j.enpol.2019.04.033 \u003c/li\u003e\n\u003cli\u003eFan, F., Lian, H., Liu, X., \u0026amp; Wang, X. (2021). Can environmental regulation promote urban green innovation Efficiency? An empirical study based on Chinese cities. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 287\u003c/em\u003e, 125060. https://doi.org/10.1016/j.jclepro.2020.125060 \u003c/li\u003e\n\u003cli\u003eFang, X. (2024). Can climate risk exposure compel companies to undergo a green transition? \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 434\u003c/em\u003e, 140310. https://doi.org/10.1016/j.jclepro.2023.140310 \u003c/li\u003e\n\u003cli\u003eFlammer, C. (2021). Corporate green bonds. \u003cem\u003eJournal of Financial Economics\u003c/em\u003e,\u003cem\u003e 142\u003c/em\u003e(2), 499-516. https://doi.org/10.1016/j.jfineco.2021.01.010 \u003c/li\u003e\n\u003cli\u003eFrank, M. R., Autor, D., Bessen, J. E., Brynjolfsson, E., Cebrian, M., Deming, D. J., Feldman, M., Groh, M., Lobo, J., \u0026amp; Moro, E. (2019). Toward understanding the impact of artificial intelligence on labor. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e,\u003cem\u003e 116\u003c/em\u003e(14), 6531-6539. https://doi.org/10.1073/pnas.1900949116 \u003c/li\u003e\n\u003cli\u003eGeng, C., \u0026amp; Wen, G. (2025). Reform of the social credit system, green innovation, and corporate resilience. \u003cem\u003eFinance Research Letters\u003c/em\u003e, 108550. https://doi.org/10.1016/j.frl.2025.108550 \u003c/li\u003e\n\u003cli\u003eGuan, T., Zheng, R., \u0026amp; Chen, A. (2026). Artificial intelligence and corporate energy consumption: The policy effects of the new-generation artificial intelligence innovation and development pilot zones. \u003cem\u003eEconomic Analysis and Policy\u003c/em\u003e,\u003cem\u003e 89\u003c/em\u003e, 148-164. https://doi.org/10.1016/j.eap.2025.11.032 \u003c/li\u003e\n\u003cli\u003eHan, S., Sun, X., \u0026amp; Zhong, S. (2025). Applications of artificial intelligence and urban innovation performance: A quasi-natural experiment based on the pilot zones for the innovative application of artificial intelligence. \u003cem\u003ePlos one\u003c/em\u003e,\u003cem\u003e 20\u003c/em\u003e(8), e0329729. https://doi.org/10.1371/journal.pone.0329729 \u003c/li\u003e\n\u003cli\u003eHerrera, A. M., \u0026amp; Minetti, R. (2007). Informed finance and technological change: Evidence from credit relationships. \u003cem\u003eJournal of Financial Economics\u003c/em\u003e,\u003cem\u003e 83\u003c/em\u003e(1), 223-269. https://doi.org/10.1016/j.jfineco.2005.12.001 \u003c/li\u003e\n\u003cli\u003eHilb, M. (2020). Toward artificial governance? The role of artificial intelligence in shaping the future of corporate governance. \u003cem\u003eJournal of Management and Governance\u003c/em\u003e,\u003cem\u003e 24\u003c/em\u003e(4), 851-870. https://doi.org/10.1007/s10997-020-09519-9 \u003c/li\u003e\n\u003cli\u003eHuang, L., \u0026amp; Lei, Z. (2021). How environmental regulation affect corporate green investment: Evidence from China. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 279\u003c/em\u003e, 123560. https://doi.org/10.1016/j.jclepro.2020.123560 \u003c/li\u003e\n\u003cli\u003eHumberd, B. K., \u0026amp; Latham, S. F. (2025). When AI becomes an agent of the firm: examining the evolution of AI in organizations through an agency theory lens. \u003cem\u003eJournal of Management Studies\u003c/em\u003e. https://doi.org/10.1111/joms.13274 \u003c/li\u003e\n\u003cli\u003eHuseynov, F., \u0026amp; Klamm, B. K. (2012). Tax avoidance, tax management and corporate social responsibility. \u003cem\u003eJournal of Corporate Finance\u003c/em\u003e,\u003cem\u003e 18\u003c/em\u003e(4), 804-827. https://doi.org/10.1016/j.jcorpfin.2012.06.005 \u003c/li\u003e\n\u003cli\u003eJensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. \u003cem\u003eThe American economic review\u003c/em\u003e,\u003cem\u003e 76\u003c/em\u003e(2), 323-329. \u003c/li\u003e\n\u003cli\u003eJia, N., Luo, X., Fang, Z., \u0026amp; Liao, C. (2024). When and how artificial intelligence augments employee creativity. \u003cem\u003eAcademy of Management Journal\u003c/em\u003e,\u003cem\u003e 67\u003c/em\u003e(1), 5-32. https://doi.org/10.5465/amj.2022.0426 \u003c/li\u003e\n\u003cli\u003eJiakui, C., Abbas, J., Najam, H., Liu, J., \u0026amp; Abbas, J. (2023). Green technological innovation, green finance, and financial development and their role in green total factor productivity: Empirical insights from China. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 382\u003c/em\u003e, 135131. https://doi.org/10.1016/j.jclepro.2022.135131 \u003c/li\u003e\n\u003cli\u003eJiang, T. (2022). Mediating effects and moderating effects in causal inference. \u003cem\u003eChina Industrial Economics\u003c/em\u003e,\u003cem\u003e 5\u003c/em\u003e(5), 100-120. https://doi.org/10.19581/j.cnki.ciejournal.2022.05.005. \u003c/li\u003e\n\u003cli\u003eJiang, W., Huang, X., Gan, L., \u0026amp; Martek, I. (2025). Dynamic pricing and green investment in the construction supply chain with incentive and consumer reference price. \u003cem\u003eEngineering, Construction and Architectural Management\u003c/em\u003e, 1-29. https://doi.org/10.1108/ecam-04-2024-0489 \u003c/li\u003e\n\u003cli\u003eKemp, R., \u0026amp; Never, B. (2017). Green transition, industrial policy, and economic development. \u003cem\u003eOxford Review of Economic Policy\u003c/em\u003e,\u003cem\u003e 33\u003c/em\u003e(1), 66-84. https://doi.org/10.1093/oxrep/grw037 \u003c/li\u003e\n\u003cli\u003eKim, S., \u0026amp; Yoon, A. (2023). Analyzing active fund managers\u0026rsquo; commitment to ESG: Evidence from the United Nations Principles for Responsible Investment. \u003cem\u003eManagement science\u003c/em\u003e,\u003cem\u003e 69\u003c/em\u003e(2), 741-758. https://doi.org/10.1287/mnsc.2022.4394 \u003c/li\u003e\n\u003cli\u003eKuang, C. E., Liu, J. Y., \u0026amp; Li, W. Y. (2024). Digital-intelligence integration empowerment and green transition of manufacturing enterprises. \u003cem\u003eContemporary Finance \u0026amp; Economics\u003c/em\u003e,\u003cem\u003e 5\u003c/em\u003e, 114-127. \u003c/li\u003e\n\u003cli\u003eLee, K. Y., Jung, H. J., \u0026amp; Kwon, Y. (2024). Boundary-spanning technology search, product component reuse, and new product innovation: Evidence from the smartphone industry. \u003cem\u003eResearch Policy\u003c/em\u003e,\u003cem\u003e 53\u003c/em\u003e(4), 104959. https://doi.org/10.1016/j.respol.2024.104959 \u003c/li\u003e\n\u003cli\u003eLi, B., Peng, X., \u0026amp; Ouyang, M. (2013). Environmental regulation, green total factor productivity and the transformation of China\u0026rsquo;s industrial development mode\u0026mdash;Analysis based on data of China\u0026rsquo;s 36 industries. \u003cem\u003eChina Ind. Econ\u003c/em\u003e,\u003cem\u003e 4\u003c/em\u003e, 56-68. https://doi.org/10.19581/j.cnki.ciejournal.2013.04.005 \u003c/li\u003e\n\u003cli\u003eLi, D., \u0026amp; Wang, P. (2025). FinTech development and corporate green total factor productivity. \u003cem\u003eInternational Review of Financial Analysis\u003c/em\u003e,\u003cem\u003e 107\u003c/em\u003e, 104567. https://doi.org/10.1016/j.irfa.2025.104567 \u003c/li\u003e\n\u003cli\u003eLi, P., Yang, D., Li, P., Ye, Z., \u0026amp; Deng, Z. (2011). A study on the green transition of Chinese industry. \u003cem\u003eChina Industrial Economics\u003c/em\u003e,\u003cem\u003e 4\u003c/em\u003e, 5-14. https://doi.org/10.19581/j.cnki.ciejournal.2011.04.001 \u003c/li\u003e\n\u003cli\u003eLi, T., Meng, X., Wang, L., \u0026amp; Gong, Y. (2025). Will China\u0026apos;s Carbon Emission Trading Regulation Really Decrease the Green Investment of Firms? A Microperspective on Corporate Governance. \u003cem\u003eBusiness strategy and the environment\u003c/em\u003e,\u003cem\u003e 34\u003c/em\u003e(2), 2222-2238. https://doi.org/10.1002/bse.4103 \u003c/li\u003e\n\u003cli\u003eLi, W., Hu, H., \u0026amp; Hong, Z. (2024). Green finance policy, ESG rating, and cost of debt\u0026mdash;\u0026mdash;Evidence from China. \u003cem\u003eInternational Review of Financial Analysis\u003c/em\u003e,\u003cem\u003e 92\u003c/em\u003e, 103051. https://doi.org/10.1016/j.irfa.2023.103051 \u003c/li\u003e\n\u003cli\u003eLin, B.-q., \u0026amp; Yang, Y.-x. (2025). Building efficiency: How the national AI innovation pilot zones enhance green energy utilization? Evidence from China. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e,\u003cem\u003e 387\u003c/em\u003e, 125945. https://doi.org/10.1016/j.jenvman.2025.125945 \u003c/li\u003e\n\u003cli\u003eLiu, C., Cui, L., \u0026amp; Li, C. (2022). Impact of environmental regulation on the green total factor productivity of dairy farming: Evidence from China. \u003cem\u003eSustainability\u003c/em\u003e,\u003cem\u003e 14\u003c/em\u003e(12), 7274. https://doi.org/10.3390/su14127274 \u003c/li\u003e\n\u003cli\u003eLiu, H., Tan, M., \u0026amp; Cao, R. (2026). The role of AI in promoting urban sustainable development: Evidence from China\u0026apos;s new generation of national AI innovative development pilot zones. \u003cem\u003eApplied Geography\u003c/em\u003e,\u003cem\u003e 188\u003c/em\u003e, 103906. https://doi.org/10.1016/j.apgeog.2026.103906 \u003c/li\u003e\n\u003cli\u003eLiu, Q., \u0026amp; Zhao, C. (2020). Tax incentives and upgrading firms\u0026rsquo; human capital. \u003cem\u003eEconomic Research Journal\u003c/em\u003e,\u003cem\u003e 55\u003c/em\u003e(4), 70-85. \u003c/li\u003e\n\u003cli\u003eLiu, Y., Wu, K., \u0026amp; Liang, X. (2024). Does low-carbon pilot policy promote corporate green total factor productivity? \u003cem\u003eEconomic Analysis and Policy\u003c/em\u003e,\u003cem\u003e 84\u003c/em\u003e, 1-24. https://doi.org/10.1016/j.eap.2024.08.015 \u003c/li\u003e\n\u003cli\u003eMaghyereh, A., Boulanouar, Z., \u0026amp; Essid, L. (2025). The dynamics of green innovation and environmental policy stringency in energy transition investments. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 487\u003c/em\u003e, 144649. https://doi.org/10.1016/j.jclepro.2024.144649 \u003c/li\u003e\n\u003cli\u003eM\u0026auml;kel\u0026auml;, H., \u0026amp; N\u0026auml;si, S. (2010). Social responsibilities of MNCs in downsizing operations: A Finnish forest sector case analysed from the stakeholder, social contract and legitimacy theory point of view. \u003cem\u003eAccounting, Auditing \u0026amp; Accountability Journal\u003c/em\u003e,\u003cem\u003e 23\u003c/em\u003e(2), 149-174. \u003c/li\u003e\n\u003cli\u003eManzaneque, M., Rojo-Ram\u0026iacute;rez, A. A., Di\u0026eacute;guez-Soto, J., \u0026amp; Mart\u0026iacute;nez-Romero, M. J. (2020). How negative aspiration performance gaps affect innovation efficiency. \u003cem\u003eSmall Business Economics\u003c/em\u003e,\u003cem\u003e 54\u003c/em\u003e(1), 209-233. https://doi.org/10.1007/s11187-018-0091-8 \u003c/li\u003e\n\u003cli\u003eMartin, G. P., Wiseman, R. M., \u0026amp; Gomez-Mejia, L. R. (2019). The interactive effect of monitoring and incentive alignment on agency costs. \u003cem\u003eJournal of Management\u003c/em\u003e,\u003cem\u003e 45\u003c/em\u003e(2), 701-727. https://doi.org/10.1177/0149206316678453 \u003c/li\u003e\n\u003cli\u003eMeng, X., Zhang, M., \u0026amp; Zhao, Y. (2024). Environmental regulation and green transition: Quasi-natural experiment from China\u0026apos;s efforts in sulfur dioxide emissions control. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 434\u003c/em\u003e, 139741. https://doi.org/10.1016/j.jclepro.2023.139741 \u003c/li\u003e\n\u003cli\u003eMijit, R., Hu, Q., Xu, J., \u0026amp; Ma, G. (2025a). Greening through AI? The impact of Artificial Intelligence Innovation and Development Pilot Zones on green innovation in China. \u003cem\u003eEnergy Economics\u003c/em\u003e, 108507. https://doi.org/10.1016/j.eneco.2025.108507 \u003c/li\u003e\n\u003cli\u003eMijit, R., Hu, Q., Xu, J., \u0026amp; Ma, G. (2025b). Greening through AI? The impact of Artificial Intelligence Innovation and Development Pilot Zones on green innovation in China. \u003cem\u003eEnergy Economics\u003c/em\u003e,\u003cem\u003e 146\u003c/em\u003e, 108507. https://doi.org/10.1016/j.eneco.2025.108507 \u003c/li\u003e\n\u003cli\u003eOh, D.-h. (2010). A global Malmquist-Luenberger productivity index. \u003cem\u003eJournal of productivity analysis\u003c/em\u003e,\u003cem\u003e 34\u003c/em\u003e(3), 183-197. https://doi.org/10.1007/s11123-010-0178-y \u003c/li\u003e\n\u003cli\u003ePang, S., Hua, G., \u0026amp; Liu, Y. (2025). Legitimacy and Opportunism: Examining Climate Risk and Green Mergers and Acquisitions in Polluting Firms. \u003cem\u003eBusiness Ethics, the Environment \u0026amp; Responsibility\u003c/em\u003e. https://doi.org/10.1111/beer.12829 \u003c/li\u003e\n\u003cli\u003eQian, S., Yang, Z., Yang, L., \u0026amp; Zhang, Y. (2025). Institutional investors\u0026rsquo; green attention and corporate greenwashing: The effectiveness of external governance. \u003cem\u003eEconomic Analysis and Policy\u003c/em\u003e. https://doi.org/10.1016/j.eap.2025.05.048 \u003c/li\u003e\n\u003cli\u003eRaisch, S., \u0026amp; Fomina, K. (2025). Combining human and artificial intelligence: Hybrid problem-solving in organizations. \u003cem\u003eAcademy of Management Review\u003c/em\u003e,\u003cem\u003e 50\u003c/em\u003e(2), 441-464. https://doi.org/10.5465/amr.2021.0421 \u003c/li\u003e\n\u003cli\u003eSacconi, L. (2006). A social contract account for CSR as an extended model of corporate governance (I): Rational bargaining and justification. \u003cem\u003eJournal of business ethics\u003c/em\u003e,\u003cem\u003e 68\u003c/em\u003e(3), 259-281. https://doi.org/10.1007/s10551-006-9014-8 \u003c/li\u003e\n\u003cli\u003eShahzad, M. F., Xu, S., An, X., Asif, M., \u0026amp; Jafri, M. A. H. (2024). Effect of stakeholder pressure on environmental performance: Do virtual CSR, green credit, environmental and social reputation matter? \u003cem\u003eJournal of Environmental Management\u003c/em\u003e,\u003cem\u003e 368\u003c/em\u003e, 122223. https://doi.org/10.1016/j.jenvman.2024.122223 \u003c/li\u003e\n\u003cli\u003eShen, G., \u0026amp; Huang, S. (2019). The impact of city-level intellectual property protection on foreign capital entry into Chinese enterprises. \u003cem\u003eFinance and trade economies\u003c/em\u003e,\u003cem\u003e 40\u003c/em\u003e, 143-157. \u003c/li\u003e\n\u003cli\u003eShen, Y., \u0026amp; Zhang, X. (2024). The impact of artificial intelligence on employment: the role of virtual agglomeration. \u003cem\u003eHumanities and Social Sciences Communications\u003c/em\u003e,\u003cem\u003e 11\u003c/em\u003e(1), 122. https://doi.org/10.1057/s41599-024-02647-9 \u003c/li\u003e\n\u003cli\u003eShi, L., Cai, Z., Ding, X., Di, R., \u0026amp; Xiao, Q. (2020). What Factors Affect the Level of Green Urbanization in the Yellow River Basin in the Context of New-Type Urbanization? \u003cem\u003eSustainability\u003c/em\u003e,\u003cem\u003e 12\u003c/em\u003e(6), 2488. https://doi.org/10.3390/su12062488 \u003c/li\u003e\n\u003cli\u003eTang, D., \u0026amp; Yang, Z. (2022). Local environmental expenditures, fiscal subsidies on environmental protection and corporate green technology innovation. \u003cem\u003ePublic Fin Res\u003c/em\u003e,\u003cem\u003e 467\u003c/em\u003e, 79-93. \u003c/li\u003e\n\u003cli\u003eTian, J., Sun, S., Cao, W., Bu, D., \u0026amp; Xue, R. (2024). Make every dollar count: The impact of green credit regulation on corporate green investment efficiency. \u003cem\u003eEnergy Economics\u003c/em\u003e,\u003cem\u003e 130\u003c/em\u003e, 107307. https://doi.org/10.1016/j.eneco.2024.107307 \u003c/li\u003e\n\u003cli\u003eTian, P., \u0026amp; Lin, B. (2017). Promoting green productivity growth for China\u0026apos;s industrial exports: evidence from a hybrid input-output model. \u003cem\u003eEnergy Policy\u003c/em\u003e,\u003cem\u003e 111\u003c/em\u003e, 394-402. https://doi.org/10.1016/j.enpol.2017.09.033 \u003c/li\u003e\n\u003cli\u003eTone, K. (2001). A slacks-based measure of efficiency in data envelopment analysis. \u003cem\u003eEuropean journal of operational research\u003c/em\u003e,\u003cem\u003e 130\u003c/em\u003e(3), 498-509. https://doi.org/10.1016/S0377-2217(99)00407-5 \u003c/li\u003e\n\u003cli\u003eTorrent‐Sellens, J., Enache‐Zegheru, M., \u0026amp; Ficapal‐Cus\u0026iacute;, P. (2025). Promoting the European Sustainable Firm: How Economic, Social, and Green Innovation and the AI‐Based Technologies Create Pathways of Social and Environmental Sustainability. \u003cem\u003eBusiness strategy and the environment\u003c/em\u003e,\u003cem\u003e 34\u003c/em\u003e(7), 9093-9119. https://doi.org/10.1002/bse.70068 \u003c/li\u003e\n\u003cli\u003eVanneste, B. S., \u0026amp; Puranam, P. (2024). Artificial intelligence, trust, and perceptions of agency. \u003cem\u003eAcademy of Management Review\u003c/em\u003e(ja), amr-2022. \u003c/li\u003e\n\u003cli\u003eWang, A., Bi, C., \u0026amp; Zhang, Y. (2025). Government accounting supervision and supply chain environmental investment. \u003cem\u003eFinance Research Letters\u003c/em\u003e,\u003cem\u003e 76\u003c/em\u003e, 106977. https://doi.org/10.1016/j.frl.2025.106977 \u003c/li\u003e\n\u003cli\u003eWang, H., Guo, T., \u0026amp; Tang, Q. (2021). The effect of national culture on corporate green proactivity. \u003cem\u003eJournal of Business Research\u003c/em\u003e,\u003cem\u003e 131\u003c/em\u003e, 140-150. https://doi.org/10.1016/j.jbusres.2021.03.023 \u003c/li\u003e\n\u003cli\u003eWang, K., Sun, K., Li, Y., \u0026amp; Xie, Q. (2025). Can digital technology break the financing dilemma of innovative SMEs? Based on the perspective of enterprise innovation. \u003cem\u003eTechnological Forecasting and Social Change\u003c/em\u003e,\u003cem\u003e 214\u003c/em\u003e, 124030. https://doi.org/10.1016/j.techfore.2025.124030 \u003c/li\u003e\n\u003cli\u003eWang, S., Wang, Y., \u0026amp; Li, C. (2024). AI-driven capital-skill complementarity: Implications for skill premiums and labor mobility. \u003cem\u003eFinance Research Letters\u003c/em\u003e,\u003cem\u003e 68\u003c/em\u003e, 106044. https://doi.org/10.1016/j.frl.2024.106044 \u003c/li\u003e\n\u003cli\u003eWang, X. Q., \u0026amp; Ning, J. H. (2020). Can mandatory social responsibility disclosure drive corporate green transition? Evidence based on green patent data of listed companies in China. \u003cem\u003eJournal of Audit \u0026amp; Economics\u003c/em\u003e,\u003cem\u003e 35\u003c/em\u003e(4), 69-77. \u003c/li\u003e\n\u003cli\u003eWu, F., Hu, H., Lin, H., \u0026amp; Ren, X. (2021). Enterprise digital transformation and capital market performance: Empirical evidence from stock liquidity. \u003cem\u003eManagement world\u003c/em\u003e,\u003cem\u003e 37\u003c/em\u003e(7), 130-144. \u003c/li\u003e\n\u003cli\u003eYan, S., Almandoz, J., \u0026amp; Ferraro, F. (2021). The impact of logic (in) compatibility: Green investing, state policy, and corporate environmental performance. \u003cem\u003eAdministrative Science Quarterly\u003c/em\u003e,\u003cem\u003e 66\u003c/em\u003e(4), 903-944. https://doi.org/10.1177/00018392211005756 \u003c/li\u003e\n\u003cli\u003eYan, Z., Yu, Y., Du, K., \u0026amp; Zhang, N. (2024). How does environmental regulation promote green technology innovation? Evidence from China\u0026apos;s total emission control policy. \u003cem\u003eEcological Economics\u003c/em\u003e,\u003cem\u003e 219\u003c/em\u003e, 108137. https://doi.org/10.1016/j.ecolecon.2024.108137 \u003c/li\u003e\n\u003cli\u003eYang, Y., \u0026amp; Chi, Y. (2023). Path selection for enterprises\u0026rsquo; green transition: Green innovation and green mergers and acquisitions. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 412\u003c/em\u003e, 137397. https://doi.org/10.1016/j.jclepro.2023.137397 \u003c/li\u003e\n\u003cli\u003eYu, F., \u0026amp; Chen, J. (2023). The impact of industrial internet platform on green innovation: Evidence from a quasi-natural experiment. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 414\u003c/em\u003e, 137645. https://doi.org/10.1016/j.jclepro.2023.137645 \u003c/li\u003e\n\u003cli\u003eYu, X. L., \u0026amp; Zhou, Y. (2023). Green Credit Policy and Green Transition of High-pollution Enterprises: A Perspective from Emission Reduction and Development. \u003cem\u003eThe Journal of Quantitative \u0026amp; Technical Economics\u003c/em\u003e,\u003cem\u003e 5\u003c/em\u003e, 1-22. https://doi.org/10.19581/j.cnki.ciejournal.2011.04.001 \u003c/li\u003e\n\u003cli\u003eYuan, B., \u0026amp; Xiang, Q. (2018). Environmental regulation, industrial innovation and green development of Chinese manufacturing: Based on an extended CDM model. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 176\u003c/em\u003e, 895-908. https://doi.org/10.1016/j.jclepro.2017.12.034 \u003c/li\u003e\n\u003cli\u003eYuxin, R., \u0026amp; Zhengchu, H. (2025). Urban Digital Intelligence Development and Corporate New Quality Productive Forces: A Quasi-natural Experiment Based on National Artificial Intelligence (AI) Innovative Development Pilot Zones. \u003cem\u003eJournal of Shanghai University of Finance and Economics\u003c/em\u003e,\u003cem\u003e 27\u003c/em\u003e(3), 34-48. https://doi.org/10.16538/j.cnki.jsufe.2025.03.003 \u003c/li\u003e\n\u003cli\u003eZhai, X., An, Y., Shi, X., \u0026amp; Liu, X. (2022). Measurement of green transition and its driving factors: Evidence from China. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 335\u003c/em\u003e, 130292. https://doi.org/10.1016/j.jclepro.2021.130292 \u003c/li\u003e\n\u003cli\u003eZhang, L., \u0026amp; Zhang, S. (2020). Technology Empowerment: The Effect of Technological Innovation on the Integrated Development of Artificial Intelligence and Industry. \u003cem\u003eFinanc. Econ\u003c/em\u003e,\u003cem\u003e 15\u003c/em\u003e. \u003c/li\u003e\n\u003cli\u003eZhang, R. (2024). Fintech-driven green economy leapfrog: an exploration of jiangsu province paths based on entropy weighted TOPSIS-coupled coordination models. \u003cem\u003eEnvironment, Development and Sustainability\u003c/em\u003e, 1-33. https://doi.org/10.1007/s10668-024-05809-6 \u003c/li\u003e\n\u003cli\u003eZhang, S., Cheng, L., Ren, Y., \u0026amp; Yao, Y. (2024). Effects of carbon emission trading system on corporate green total factor productivity: Does environmental regulation play a role of green blessing? \u003cem\u003eEnvironmental research\u003c/em\u003e,\u003cem\u003e 248\u003c/em\u003e, 118295. https://doi.org/10.1016/j.envres.2024.118295 \u003c/li\u003e\n\u003cli\u003eZhang, S., Dong, R., Jiang, J., Yang, S., Cifuentes-Faura, J., Peng, S., \u0026amp; Feng, Y. (2024). Whether the green credit policy effectively promote green transition of enterprises in China? Empirical analysis and mechanism verification. \u003cem\u003eEnvironmental research\u003c/em\u003e,\u003cem\u003e 244\u003c/em\u003e, 117910. https://doi.org/10.1016/j.envres.2023.117910 \u003c/li\u003e\n\u003cli\u003eZhang, W., \u0026amp; Sun, C. (2023). How does outward foreign direct investment affect enterprise green transition? Evidence from China. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 428\u003c/em\u003e, 139331. https://doi.org/10.1016/j.jclepro.2023.139331 \u003c/li\u003e\n\u003cli\u003eZhang, X., Jin, Y., \u0026amp; Shen, C. (2020). Manufacturers\u0026rsquo; green investment in a competitive market with a common retailer. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 276\u003c/em\u003e, 123164. https://doi.org/10.1016/j.jclepro.2020.123164 \u003c/li\u003e\n\u003cli\u003eZhao, M., Gao, Y., Feng, J., \u0026amp; Sun, W. (2024). Effect test of China\u0026rsquo;s outward foreign direct investment reverse technology spillover on green total factor production under cultivation of new quality productivity forces. \u003cem\u003eBulletin of Chinese Academy of Sciences (Chinese Version)\u003c/em\u003e,\u003cem\u003e 40\u003c/em\u003e(2), 350-361. \u003c/li\u003e\n\u003cli\u003eZhao, X., Nakonieczny, J., Jabeen, F., Shahzad, U., \u0026amp; Jia, W. (2022). Does green innovation induce green total factor productivity? Novel findings from Chinese city level data. \u003cem\u003eTechnological Forecasting and Social Change\u003c/em\u003e,\u003cem\u003e 185\u003c/em\u003e, 122021. https://doi.org/10.1016/j.techfore.2022.122021 \u003c/li\u003e\n\u003cli\u003eZhou, Z., Nie, L., Ji, H., Zeng, H., \u0026amp; Chen, X. (2020). Does a firm\u0026rsquo;s low-carbon awareness promote low-carbon behaviors? Empirical evidence from China. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e,\u003cem\u003e 244\u003c/em\u003e, 118903. https://doi.org/10.1016/j.jclepro.2019.118903 \u003c/li\u003e\n\u003cli\u003eZhu, L., Li, T., Wang, C., \u0026amp; Huang, J. (2025). Corporate ESG performance and green innovation: moderating effect of shareholder activism. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e,\u003cem\u003e 383\u003c/em\u003e, 125413. https://doi.org/10.1016/j.jenvman.2025.125413 \u003c/li\u003e\n\u003cli\u003eZhu, X., Liu, R., \u0026amp; Chen, J. (2023). Corporate environmental investment and supply chain financing: The moderating role of environmental innovation. \u003cem\u003eBusiness strategy and the environment\u003c/em\u003e,\u003cem\u003e 32\u003c/em\u003e(4), 1559-1581. https://doi.org/10.1002/bse.3205 \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"humanities-and-social-sciences-communications","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"palcomms","sideBox":"Learn more about [Humanities \u0026 Social Sciences Communications](http://www.nature.com/palcomms/)","snPcode":"41599","submissionUrl":"https://submission.springernature.com/new-submission/41599/3","title":"Humanities and Social Sciences Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"artificial intelligence, corporate green transition, green investment, credit access, green innovation, agency cost","lastPublishedDoi":"10.21203/rs.3.rs-8993678/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8993678/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn the context of the synergistic promotion of the digital economy and green transition, the impact of AI development on enterprises' green transition is receiving increasing attention. Based on data on A-share listed companies in Shanghai and Shenzhen from 2010 to 2023, this paper constructs a quasi-natural experiment using the AI pilot test zone policy and employs a double-difference model to systematically identify the impact of AI policy on enterprises' green transition behaviour. It is found that the construction of AI pilot zone significantly promotes the improvement of enterprises' green transition level, and this conclusion still holds under multiple robustness tests. Mechanism analyses show that AI policies primarily promote corporate green transition through multiple pathways, including increasing corporate green investment, strengthening credit availability, improving the efficiency of green innovation, and reducing agency costs. The heterogeneity test based on the TOE analysis framework further reveals that the above policy effects are more significant in firms with a higher degree of digital transformation, a sufficient reserve of high-tech talent, better internal governance and better ESG performance; meanwhile, the policy promotion effect is further enhanced in regions with higher intensity of intellectual property rights (IPR) protection systems and environmental regulations. This paper reveals the mechanisms by which AI policies promote green transition at the micro-firm level, providing new empirical evidence for understanding AI-driven green development.\u003c/p\u003e","manuscriptTitle":"Artificial Intelligence and Corporate Green Transition--Evidence From China's AI Innovation Development Pilot Zones","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-20 10:48:16","doi":"10.21203/rs.3.rs-8993678/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-15T14:06:45+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-31T03:26:56+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-26T06:33:04+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"264140651908173042718161874774909828886","date":"2026-03-26T06:23:16+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"280317628732291079775891396748991175931","date":"2026-03-20T08:50:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"49092607055760951120899940198677002072","date":"2026-03-18T14:44:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"4968680184810645920653262717288848915","date":"2026-03-18T09:45:34+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-17T14:06:35+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-17T11:59:48+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-11T09:52:52+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-09T19:00:49+00:00","index":"","fulltext":""},{"type":"submitted","content":"Humanities and Social Sciences Communications","date":"2026-03-05T19:35:18+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"humanities-and-social-sciences-communications","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"palcomms","sideBox":"Learn more about [Humanities \u0026 Social Sciences Communications](http://www.nature.com/palcomms/)","snPcode":"41599","submissionUrl":"https://submission.springernature.com/new-submission/41599/3","title":"Humanities and Social Sciences Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"0ebb66c2-eb21-49c7-a468-8f55ab85ae52","owner":[],"postedDate":"March 20th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":64734357,"name":"Business and commerce/Business and management"},{"id":64734358,"name":"Social science/Business and management"},{"id":64734359,"name":"Earth and environmental sciences/Environmental social sciences"}],"tags":[],"updatedAt":"2026-04-15T14:11:10+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-20 10:48:16","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8993678","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8993678","identity":"rs-8993678","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}