AI Disclosure as an Institutional Investment Catalyst: Evidence from China

AI Disclosure as an Institutional Investment Catalyst: Evidence from China | 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 Research Article AI Disclosure as an Institutional Investment Catalyst: Evidence from China Feng zhao, Dandan Nan, Yao Wu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7134593/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Artificial intelligence (AI) is a significant driver propelling the new wave of technological revolution and industrial transformation, exerting a profound influence across various sectors. This paper examines the relationship between AI information disclosure and changes in institutional investor shareholdings, utilizing data from Chinese A-share companies listed on the Shanghai and Shenzhen stock exchanges between 2007 and 2023. The results indicate that enhanced AI information disclosure is associated with increased institutional investor ownership. This conclusion remains valid following robustness tests that include substituting dependent and independent variables, altering clustering levels, excluding municipal samples, and employing quantile regression. Heterogeneity analyses further reveal that the effect is notably stronger in samples post-2011, within manufacturing firms, in environments with higher competition, and among firms with directors possessing a strong technological background. The mechanism analysis reveals that increased AI information disclosure prompts institutional investors to conduct on-site visits, enhancing information transparency and bolstering innovation capabilities. Further, heightened AI information disclosure is particularly significant in encouraging shareholdings by long-term and resilient institutional investors. This paper presents the first micro-level evidence on the impact of AI disclosure on institutional investor shareholdings, offering theoretical and practical implications for refining investment strategies of institutional investors in the capital market. JEL Classification G10 G32 artificial intelligence information disclosure institutional investor shareholding on-site visits information transparency innovation Figures Figure 1 1. Introduction Since the 18th Party Congress, China has placed significant emphasis on the advancement of artificial intelligence. In 2015, the initiative “Made in China 2025” explicitly set forth the objective of capitalizing on the opportunity presented by the new wave of technological revolution, accelerating the implementation of development strategies, and establishing a powerful manufacturing nation. Consequently, artificial intelligence has advanced into a phase of swift development, with the artificial intelligence market experiencing continuous expansion. The 2017 “New Generation Artificial Intelligence Development Plan” constructs the strategic goal of “2020 synchronization-2025 breakthrough-2030 leadership”. Starting from 2019, a total of 18 pilot zones for the innovation and development of artificial intelligence have been set up in three batches. The 2025 government work report explicitly proposes to continue advancing the “artificial intelligence +” initiative, and to promote the integration of artificial intelligence with the real economy. Driven by national policies, artificial intelligence has been widely used in industries such as medicine (Iacucci et al., 2025), finance (Du et al., 2025), education (Wang et al., 2025), and food (Nath et al., 2024), etc. The latest IDC 1 report in 2025 indicates that China will continue to lead the Asia-Pacific region’s artificial intelligence development, with total artificial intelligence investment expected to exceed $ 100 billion by 2028, at a five-year CAGR of 35.2 percent. Artificial intelligence technology is profoundly reshaping the architecture of the economy and society, encompassing various key branches such as machine learning (ML), deep learning (DL), and natural language processing (NLP)(Ahmed et al., 2022). Machine learning, deep learning, and other algorithms can help companies to predict financial risks (Veganzones et al., 2025; Lin et al., 2025), and natural language processing can extract information from annual reports that is overlooked by investors, and this potential information can provide investors with investment advice (Cohen et al., 2020). Prior study focuses on the impact of artificial intelligence applications on corporate innovation (Bahoo et al., 2023; Zhong and Song, 2025), macroeconomics (Alonso et al., 2022; Das et al., 2025), and stock forecasting (Chopra et al., 2024). However, there is a question of whether artificial intelligence will affect institutional investor ownership. Against the backdrop of the country’s great attention to artificial intelligence, companies have also begun to use artificial intelligence technology to manage their operations, and an increasing number have chosen to include the application of artificial intelligence in their annual reports, such as investments in artificial intelligence and artificial intelligence patents. As a professional investment entity in the capital market, the most significant difference between institutional investors and individual investors is that they focus more on the long-term operation and development of enterprises, and the increase in their shareholding ratio has a positive impact in multiple dimensions. In a way, the disclosure of artificial intelligence information reflects the enterprise’s planning and commitment to long-term development. This information often attracts the close attention of institutional investors and then has an important impact on their shareholding decisions. Therefore, it is of great significance to explore the intrinsic relationship between corporate artificial intelligence information disclosure and institutional investor ownership. Based on data from Chinese A-share listed companies on the Shanghai and Shenzhen stock exchanges from 2007 to 2023, this paper investigates the impact and mechanism of artificial intelligence information disclosure on changes in institutional investor shareholdings. The reasons for selecting China as the research subject are threefold. Firstly, over the past decade, China’s artificial intelligence sector has experienced rapid development. The “Research Report on China’s Artificial Intelligence Regional Competitiveness” indicates that the scale of China’s artificial intelligence industry has expanded from 10 billion in 2015 to 700 billion in 2024, sustaining a growth rate of over 20% annually for consecutive years. What’s more, IBM’s 2 “Global Artificial Intelligence Adoption Index 2023” notes that approximately 50% of Chinese companies are actively implementing artificial intelligence solutions, encompassing areas such as smart healthcare and smart transportation, thus offering a substantial sample of artificial intelligence applications. Secondly, in 2004, the State Council issued “Several Opinions on Promoting the Reform, Opening-up, and Stable Development of the Capital Market”. By 2006, the capital market had experienced rapid expansion, propelling China’s institutional investors into a phase of rapid development and establishing a pattern where fund companies, finance companies, insurance companies, banks, QFIIs, and other diverse entities coexist (Jia and Che, 2025). Moreover, domestic financial databases such as Wind and CSMAR offer detailed data on institutional investor shareholdings, with a very high level of data availability, providing a rich data source for this paper. Thirdly, China has become the world’s second-largest economy, following the United States, and is also the largest emerging economy. Thus, examining the Chinese model can offer insights not only for governments seeking to enhance capital market laws and regulations but also provide empirical evidence to other emerging economies on managing investment shifts due to artificial intelligence (Song, 2024). This paper examines the Management Discussion and Analysis (MD&A) section of corporate annual reports, tallying the frequency of AI-related terminology to create an artificial intelligence information disclosure index. Baseline findings suggest that enhanced artificial intelligence disclosure encourages institutional investors to boost their stakes. This conclusion remains valid after undergoing various robustness tests, including the substitution of dependent and independent variables, increasing clustering levels, excluding municipal samples, and employing alternative estimation methods. To tackle potential endogeneity concerns, the study confirms the reliability of its findings using four distinct approaches: the instrumental variable (IV) method with two-stage least squares (2SLS), the Heckman two-step test, propensity score matching (PSM), and the inclusion of omitted variables. Subsequently, the paper investigates the economic mechanisms by which artificial intelligence disclosure amplifies institutional investor ownership. Corporate artificial intelligence disclosure aids institutional investors in conducting site visits, enhances information transparency, and fosters innovation. Heterogeneity analysis indicates that the positive impact of artificial intelligence disclosure on institutional investor shareholdings is more pronounced in post-2011 samples, among manufacturing firms, in highly competitive environments, and for companies with directors possessing a strong technological background. Further analysis demonstrates that the influence of changes in artificial intelligence disclosure on institutional investor shareholdings is more significant for long-term and stress-resistant institutional investors compared to their short-term and stress-sensitive counterparts, highlighting the varying decision-making processes among different investment entities. There are three main contributions in this study. Firstly, prior studies examine firm characteristics (Gaver and Gaver, 1993; Hong et al., 2008), corporate governance level (Chung and Zhang, 2011), market performance (Deb, 2018), macro uncertainty (Huang et al., 2022; Laksmana et al. 2023) and corporate social responsibility (Marshall et al., 2022; Yahia et al., 2023) on institutional investor ownership. Some scholars have also examined the impact of disclosure on institutional investor shareholdings from the perspective of corporate disclosure quality (Cheng et al., 2020). We concentrate on artificial intelligence, thereby broadening the body of research on institutional investor shareholding. Secondly, we enrich the research on the economic consequences of artificial intelligence. Previous research mostly investigates the impact of artificial intelligence on supply chain (Wu et al., 2025; Yang et al., 2025), productivity (Sullivan and Wamba, 2024; Sun et al., 2024), investment efficiency (Wang and Chen, 2025 ), employment (Saba and Ngepah, 2024), population aging (Wen et al., 2025), environment (Wang et al., 2024), energy (Gao and Wang, 2025 ), ESG (Zhou et al., 2025), green finance (Wang et al., 2025) and green transition (Li et al., 2025). However, there is limited research on the impact of artificial intelligence on institutional investors within the capital market. Consequently, our study aims to address this gap concerning the role of artificial intelligence in the capital market. Thirdly, we reveal that the disclosure of artificial intelligence information contributes to changes in institutional investor shareholdings through three distinct economic mechanisms: institutional investor site visits, information transparency, and corporate innovation. Our study elaborates on the mechanism of action by which artificial intelligence influences institutional investor shareholdings. The remainder of the paper is organized as follows. Section 2 reviews the related literature. Section 3 is the theoretical analysis and hypothesis formulation. Section 4 describes our research design. Section 5 reports our empirical results. Section 6 presents the additional analyses, and Section 7 concludes. 2. Literature review 2.1 Artificial intelligence 2.1.1 Factors affecting artificial intelligence Artificial intelligence is influenced by a variety of factors, and scholars have analyzed it from micro, industry, and macro perspectives. At the micro level, data resources, human capital, and technological capabilities all influence the development of artificial intelligence(Pinski et al., 2024). Rich data resources provide ample material for training artificial intelligence models. Professionals can integrate AI technology with the company’s core business, and advanced technology aids in the design and optimization of artificial intelligence models. Artificial intelligence represents a long-term investment, and the executive team plays a crucial role in the development strategy and the extent of artificial intelligence application (Li and Shao, 2023). On the one hand, the executive team’s artificial intelligence literacy underscores the significance of AI within the organization and motivates the application of artificial intelligence technology to actual business operations (Pinski et al., 2024). On the other hand, executives with a future-oriented perspective can incorporate artificial intelligence into the company’s long-term development strategy (Li et al., 2021). In addition, Liu et al. (2025) also found that larger firms with higher firm value are more inclined to apply artificial intelligence. At the industry level, within the financial sector, artificial intelligence undertakes tasks such as risk assessment, stock trading, and loan provision, among others. It has changed the way the financial industry provides services to its customers, which can not only improve the efficiency of internal processes, maintain cybersecurity, and reduce financial risks, but can also lead to more personalized products and services (Polireddi, 2024). In the manufacturing industry, the use of artificial intelligence enables real-time monitoring of the production floor, allowing for the identification and prompt resolution of problems. Robots, as automation-oriented applications of artificial intelligence, can replace humans in challenging and hazardous tasks. This not only enhances productivity but also ensures the safety of employees (Lee et al., 2022). Furthermore, industries with a high level of technological intensity are more inclined to invest significantly in artificial intelligence. This includes sectors such as information transmission, software and information technology services, and scientific research and technology services, which possess robust research and development (R&D) capabilities in digital technology. These industries tend to favor the use of cloud computing, big data, and artificial intelligence for managing their companies (Yao et al., 2024). Macro policy factors: Since 2013, China has released several documents on artificial intelligence, including “The New Generation Artificial Intelligence Development Plan,” “The Guidelines for the Construction of the National New Generation Artificial Intelligence Standard System,” and “The Guidelines for the Construction of the National Comprehensive Standardization System for the Artificial Intelligence Industry (Version 2024)” etc. Roberts et al. (2021) analyzed the nature of China’s artificial intelligence policy and the context in which it has emerged, suggesting that there are opportunities for China to develop in artificial intelligence areas such as automation, which will strongly contribute to the growth of the Chinese economy. Wang et al. (2025) analyzed China’s artificial intelligence policy using Structural Thematic Modeling (STM) and indicated that China has paid increasing attention to the development of artificial intelligence, focusing on applications in four areas: industrial applications, technical standards, talent training and scientific research. The proposed artificial intelligence policy will encourage enterprises to increase their capital investment in artificial intelligence and patent applications, but weak intellectual property rights will inhibit their investment in artificial intelligence innovation and reduce the incentives for patent development. Therefore, it is also crucial to improve the intellectual property protection system. China has gradually improved its IPR protection system since 2008, and has issued relevant policy documents, such as “the Outline of National Intellectual Property Strategy”, “Measures for the Evaluation of National Intellectual Property Pilot and Demonstration Cities (Districts)”, “and Outline for the Construction of a Strong Intellectual Property Rights Country (2021–2035)”, etc. Han et al. (2025) employed a time-varying difference model to study the relationship between the intellectual property model city policy and the development of enterprises’ artificial intelligence which suggests that the policy significantly improves the development level of artificial intelligence. 2.1.2 Economic impact of artificial intelligence The existing literature primarily addresses the economic implications of artificial intelligence at the enterprise operational level, the macroeconomic level, and the energy and environmental levels. In terms of enterprise operations, artificial intelligence has been rapidly applied to various fields of the economy and society, reshaping the production and operational modes of enterprises (Culot et al., 2024). Supply chain management: Artificial intelligence technology can provide supply chain services for manufacturing enterprises, aiding in the enhancement of decision-making efficiency and cost reduction through the optimization of inventory management, logistics, and distribution(Berente et al., 2021), and improving supply chain performance (Belhadi et al., 2022). Productivity: Arithmetic power, serving as the cornerstone of intelligent transformation in the era of artificial intelligence, empowers enterprises to extract value and reconfigure the allocation of production factors, thereby enhancing enterprise productivity (Ding and Gao, 2025). The supply chain digitalization transformation policy further enhances the role of AI technology in promoting the productivity of enterprises (Sun et al., 2024). Industrial robotics, as a significant application of artificial intelligence, can enhance the total factor productivity of enterprises and drive their high-quality development (Duan et al., 2023). Czarnitzki et al. (2023) constructed a firm-level Cobb-Douglas production function to further verify the positive relationship between artificial intelligence technology and firm productivity. Labor Income Level: Existing literature indicates that the impact of artificial intelligence on labor income varies by industry and is influenced by policy. In the service sector, artificial intelligence reduces the number of less-educated workers and front-line employees, thereby decreasing the firms’ share of labor income (Chen et al., 2024). And in the financial sector, Chang (2025) found that artificial intelligence can effectively increase income levels, especially on the East Coast and market-driven regions. Fan et al. (2025) employed industry data to find that next-generation artificial intelligence pilot zone policies increase firms’ share of labor income through skill demand effects and skill premium effects, which highlights the impact of artificial intelligence policies on labor income. At the macroeconomic level, Chishti et al. (2024) considered artificial intelligence as a new driver of the economic cycle. It is documented that AI may cause economic volatility in the short term, yet it has the potential to stabilize the economy in the long term. In terms of economic growth, Saba and Ngepah (2024) examined the impact of artificial intelligence on the economies of the BRICS countries using the Cobb-Douglas production function and the Solow-Swan model. Their results indicated that artificial intelligence has a significant positive impact on the economic growth of the BRICS countries, which is evident in both the short and long term, though it is more pronounced in the short term. Alonso et al. (2022) stated that artificial intelligence may exacerbate the economic divide between developing and developed economies. The economic gap between these economies exists because developed nations utilize more robots, have higher starting wages, greater robot productivity, and experience significantly higher GDP growth rates compared to developing economies. In addition, based on the extended Solow-Swan growth model, Skare et al. (2024) treated artificial intelligence as a form of capital that can replace or supplement traditional forms of labor, and found a significant positive correlation between artificial intelligence capital stock and wealth inequality, which can be explained by the technology substitution effect, that is, the growth of income of high-skilled labor is relatively fast and low-skilled labor income growth relatively slow, leading to wealth-income inequality. However, the increase in capital investment can improve productivity and thus lead to overall economic growth, but it does not change the phenomenon of wealth-income inequality. In the study of artificial intelligence and unemployment, Nguyen and Vo (2022) revealed that artificial intelligence can exacerbate unemployment; however, when inflation reaches a certain threshold, unemployment rates tend to decrease. This suggests that artificial intelligence may help to mitigate unemployment when inflation is at its anticipated level. As for Energy and Environment: In recent years, artificial intelligence has been extensively applied in the fields of energy and the environment, facilitating the transition of the economy towards a low-carbon energy structure (Huang et al., 2025). Artificial intelligence can significantly reduce energy consumption by encouraging companies to innovate and drive digital transformation (Fu et al., 2024), improve energy and resource efficiency (Li et al., 2023; Li et al., 2025), and reduce ecological footprints (Wang et al., 2024). In the area of corporate environmental performance, the application of artificial intelligence significantly improves the quality of corporate environmental disclosure (Wu et al., 2025), promotes the growth of environmental performance (Wang et al., 2024), and can effectively transform the growth of environmental performance into the enhancement of corporate reputation and market value (Liu et al., 2025). In the field of carbon emissions, artificial intelligence will reduce carbon emissions by improving production efficiency, optimizing resource allocation, and promoting technological progress (Luqman et al., 2024). Based on cross-country data, Wang et al. (2024) further pointed out that there exists a threshold effect of trade openness on the suppression of carbon emissions by artificial intelligence. When trade openness is below this threshold, the impact of artificial intelligence on carbon emissions is negligible. Conversely, when trade openness surpasses the threshold, artificial intelligence significantly reduces carbon emissions. Furthermore, artificial intelligence can mitigate carbon inequality between nations, and its inhibitory effect on carbon inequality becomes more pronounced as the severity of the inequality increases (Zhao et al., 2024). 2.2 Institutional investor shareholdings 2.2.1 Factors affecting institutional investors shareholdings It is well documented that institutional investors consider three levels of factors when making investment decisions: micro, industry, and macro. At the micro level, Cheng et al.(2020) indicated that during market downturns, the quality of corporate disclosure exerts a greater positive influence on institutional investor ownership. Institutional investors prefer to invest in listed companies that exhibit faster growth and better profitability (Lin et al., 2014). Because such companies have more resources to develop core products, which provides conditions for them to seize market share. Institutional investors assess a company’s growth to ascertain whether its stock price is overvalued or undervalued. If they find the stock to be undervalued, they may purchase shares with the expectation of earning a return as the company expands in the future. Gaver and Gaver (1993) constructed an indicator to assess a company’s growth potential, based on six variables, and found that companies on the ascent typically exhibit lower debt-to-equity ratios and dividend yields. These characteristics could potentially draw the attention of institutional investors. Companies with robust profitability also capture investors’ interest, as these firms often boast stable cash flows and a more secure financial standing. Hong et al. (2008) measured the profitability of a company in terms of ROE, and they verified that it has a significant positive impact on stock prices, thus attracting investors. Corporate Social Responsibility (CSR) Dimension: As the green economy develops, CSR has increasingly become an indicator for evaluating the investment value of companies. Those with strong social responsibility are often more appealing to institutional investors for investment (Lyssimachou and Bilinski, 2023). Li and Lu (2016) even regarded corporate environmental capital expenditure as part of CSR and indicated that environmental capital expenditure promotes the increase of institutional investors’ shareholding. In addition, the governance level of listed companies determines the efficiency of property allocation and the operation level of enterprises, and institutional investors prefer to hold shares of companies with a good governance structure, and their shareholding ratio will increase with the increase of the quality of corporate governance (Chung and Zhang, 2011). Industry Level: The “dual-carbon” goal is further propelling the development of China’s new energy industry, aligning with the global green economy’s development trend. The new energy industry encompasses a broad spectrum of sectors, including solar and wind energy, among others. Institutional investors are optimistic about the energy transition, which is expected to encourage them to invest in stocks and benefit from the dividends of this transition (Persad et al., 2024). In the healthcare sector, the growing aging population is driving the expansion of the healthcare industry. Simultaneously, ongoing innovation and breakthroughs in medical technology are providing significant momentum to the industry’s development, offering long-term and stable investment opportunities for institutional investors (Roller,2021). Additionally, the technology industry boasts a high level of innovation and growth potential, which can inject continuous vitality into economic development, thereby attracting institutional investors to invest. At the macro level, during periods of high political uncertainty, institutional investors utilize the available information to assess the impact of a political event on the market. Consequently, they decide whether to increase or decrease their stock holdings. Francis et al. (2021) stated that institutional investors tend to significantly reduce their common stock positions when political uncertainty is high. Nevertheless, if firms voluntarily disclose information pertaining to political expenditures, this can mitigate the negative impact of political uncertainty on institutional investor ownership (Goh et al., 2020). From the perspective of economic policy uncertainty (EPU), Hao and Li (2024) argued that this uncertainty can render future cash flows unstable, causing institutional investors to adopt a more cautious approach to investing in domestic stocks, thereby reducing their holdings. Wang et al. (2024) further discussed the impact of EPU on investor categories and found that long-term institutional investors react more positively to EPU because they believe that it is a short-term economic volatility behavior that will not affect the long-term value of firms; whereas, short-term institutional investors reduce their stock holdings to cope with market risks. From a global perspective, global volatility also affects the behavior of institutional investors. During periods of high volatility, institutional investors tend to decrease their stock allocations (Kacperczyk et al., 2022). In addition, geopolitical risk (GPR) is also a major factor affecting institutional investor ownership. Choudhury (2025) investigated the impact of GPR on U.S. investors through an extended five-factor model and analyzed the relationship between the GPR factor and the portfolio return, which provides empirical evidence for the adjustment of investment strategies by institutional investors. Furthermore, Fiorillo et al. (2024) demonstrated that GPR enhances institutional investors’ investment preference for green bonds. 2.2.2 Economic impact of institutional investors shareholdings Meanwhile, institutional investors exert a constructive influence on corporate governance, climate risk, and ESG. In the realm of corporate governance, institutional investors, as professional participants in the capital market, directly shape the governance framework through their ownership stakes. Appel et al. (2016) challenged the view that passive investors lack governance incentives, revealing that by concentrating their voting power, they significantly improve the corporate governance structure and bolster the long-term performance of firms. McCahery et al. (2016) employed a questionnaire to find that institutional investors are able to influence corporate governance through behind-the-scenes intervention and exit mechanisms. Aggarwal et al. (2011) suggested a positive correlation between institutional investor shareholdings and company valuations, indicating that institutional investor shareholding not only affects corporate governance but also affects company value and board decisions. Moreover, when portfolio company value appreciates, institutional investors can also receive financial incentives such as management fees, which heighten their motivation to participate in corporate governance (Lewellen and Lewellen, 2022). Concerning climate risk, climate risk poses potential financial risks for institutional investors (Krueger et al., 2019). Site visits, as a crucial engagement channel between companies and investors, inhibit managerial incentives to conceal climate risk information, enhance the quality of the information environment, and compel management to disclose more information about climate change (Song and Xian, 2024). Flammer et al. (2021) argued that institutional investors, especially long-term institutional investors, play an important role in promoting voluntary disclosure of climate risk information. Li et al. (2023) employed the event study method to uncover that climate risk information disclosure can send positive signals to institutional investors, showing significant positive abnormal returns in the samples with high investor attention, which indicates that the market’s attitude towards climate risk information disclosure is generally positive. In addition, institutional investors mitigate corporate climate risk exposure by promoting green innovation capabilities (Jin et al., 2024). Institutional investors significantly accelerate the development of corporate ESG. Huang et al. (2025) indicated that institutional investors ask ESG-related questions during site visits, which motivates companies to enhance the disclosure of non-financial information on ESG. Analysis of cross-country data by Dyck et al. (2019) confirmed that institutional investor ownership is positively associated with corporate environmental and social (E&S) performance, demonstrating that E&S decisions are not only influenced by managers but also driven by investors. Crucially, Zhang et al. (2025) reported that institutional investor ownership deters corporate greenwash while improving ESG performance, which in turn reduces operational risk and increases firm value. After further categorizing institutional investors, Liu et al. (2023) and Giordino et al. (2025) suggested a positive association between long-term institutional investors and ESG performance. 3. Theoretical analysis and hypothesis development 3.1 Artificial intelligence information disclosure and institutional investor ownership From the perspective of information economics, Stiglitz (2000) indicated that imperfect information and information friction costs trigger information asymmetry, which distorts resource allocation and contributes to market inefficiency. The advent of the Fourth Industrial Revolution presents artificial intelligence as a transformative solution to this dilemma. AI technology can effectively break down “data silos” and reduce the information friction costs of enterprises by enhancing their information acquisition and processing capabilities (Goldfarb and Tucker, 2019; He et al., 2025). Boot et al. (2021) indicated that the AI-powered analysis of unstructured data reveals a significant scale advantage in the financial services sector, effectively bridging the information gaps between financing counterparts. At the corporate level, Begenau et al. (2018) argued that compared to small enterprises, large enterprises attain greater efficiency in AI technology. They can also enter the capital market with lower thresholds, thereby attracting institutional investors. Moreover, companies that disclose more information can enable external investors to assess the firm’s value, reducing information asymmetry (Chung et al., 2015). From the Resource-Based View (RBV), Wernerfelt (1984) and Barney (1991) pointed out that resources with value, scarcity, inimitability, and non-substitutability (VRIN)‌ are the prerequisites for a company to have a competitive advantage (Liu et al., 2025a). AI is precisely a VRIN resource (Bag et al., 2021; Li et al., 2025), which belongs to the category of intangible assets (Xia et al., 2025). Among them, value is reflected in the role of artificial intelligence in improving the operational efficiency of enterprises (Slimani et al., 2025); scarcity comes from the three elements of algorithms, arithmetic power, and data; inimitability refers to the complexity of artificial intelligence technology, which is difficult to be copied by competitors; and non-substitutability is reflected in the processing of data and optimization of algorithms. Management can formulate business strategies and participate in corporate governance based on artificial intelligence (Qu and Jing, 2025 ), thus releasing positive signals of long-term stable development to the market, which makes it easier for firms applying artificial intelligence technology to gain the favor of institutional investors and increase their shareholdings. Therefore, we hypothesize as follows: H1. Corporate artificial intelligence information disclosure will prompt institutional investors to increase their stock holdings. 3.2 Artificial intelligence information disclosure, institutional investor site visits and institutional investor ownership The opacity of artificial intelligence’s algorithmic black boxes has exacerbated social trust complexities, particularly undermining institutional investors’ confidence in AI-related information disclosure in their annual reports. To explore the use of artificial intelligence and verify the authenticity of artificial intelligence technology, institutional investors increasingly employ site visits as a due diligence mechanism, which can not only observe the company’s R&D circumstance, grasp the production status and development plans (Wu et al., 2025), but can also obtain the information not disclosed in the annual reports (Zhao et al., 2023), understand the company’s prospects and business risk exposure (Jiang and Yuan, 2018), and then make investment decisions to influence the operation of the capital market (Belev, 2003; Saci & Jasimuddin, 2021; Xiao, 2023; Zhao et al., 2023). Furthermore, institutional investors frequently convene meetings following site visits, during which they engage in detailed discussions with management. These meetings also furnish institutional investors with more reliable information, influencing their decisions on stock ownership (Zhang & Ye, 2023). Furthermore, following the site visits, company management will proactively enhance the quality of information disclosure to prevent an increase in the degree of information asymmetry between individual and institutional investors, thereby improving stock liquidity (Balakrishnan et al., 2014) and promoting the increase of institutional holdings. Then, we hypothesize as follows: H2. Artificial intelligence information disclosure can promote institutional investor shareholdings by increasing the frequency of institutional investors’ field visits. 3.3 Artificial intelligence information disclosure, information transparency and institutional investor ownership According to the Attention-Based View (Ocasio, 1997), corporate disclosure of artificial intelligence information can be viewed as a conscious attention-guiding behavior. This selective disclosure strategically directs organizational focus toward AI-related matters, thereby influencing the salience allocated to such information within the firm and ultimately reshaping the attention distribution pattern (Joseph and Wilson, 2018). As an information intermediary, the media, with its unique perspective on corporate information, presents a dual effect: it not only strengthens the transparency of corporate information, thereby increasing social attention, but also plays a supervisory role in corporate governance (Raimondo, 2019). By reporting positive news, such as the advantages that AI applications bring to a company’s operations, the media can motivate companies to focus more on the research and development of artificial intelligence. Conversely, if the press highlights the disadvantages of AI technology, such as data breaches, this can compel companies to disclose more information about their artificial intelligence initiatives to rebuild their corporate image and salvage their reputation. Whether it is a positive or negative report, the press can shape the information environment of a firm (Bushee et al., 2010) and increase information transparency by packaging and disseminating information. Institutional investors are able to get more comprehensive information through these news reports, which is beneficial to their investment decisions. Analysts, as professional information decoders (Rees et al., 2015), are an important part of the capital market, which can reduce the degree of information asymmetry between the company and investors, and encourage investors to invest. Among the various types of investors, institutional investors are closely linked to analysts. In the context of the increasing complexity of the capital market, institutional investors are increasingly relying on analysts’ interpretation of data (Zou et al., 2025). The advent of artificial intelligence has elevated this dependency to a new level. With the assistance of AI technology, analysts can more accurately evaluate a company’s profitability, management, and sustainable growth, and then compile these findings into professional analysis reports. Consequently, institutional investors can utilize these reports to assess a company’s fundamental status and subsequently determine whether to increase or decrease their shareholdings (Kong et al., 2021). This synergy between the media and analysts enhances the transparency of corporate information and reduces the cost of obtaining corporate information for institutional investors (Armstrong et al., 2011), thus providing strong support for their investment decisions. Thus, we hypothesize as follows: H3. Corporate artificial intelligence information disclosure can increase institutional investor ownership by enhancing information transparency. 3.4 Artificial intelligence information disclosure, innovation and institutional investor ownership In the digital age, innovation is influenced by factors such as technological change, economic development, and human capital (Haefner et al., 2021). The nature of innovation is a learnable process of “generating new ideas - creative thinking - commercialization of services” (Bahoo et al., 2023), which provides long-term value to investors through the creation of competitive advantage and the ability to serve new markets (Sakaki & Jory, 2019). Corporations’ growth prospects are a factor influencing institutional investor shareholdings, and innovation capability, as a major assessment indicator of firms’ growth, not only reflects firms’ ability to perceive technological opportunities, but also reflects the potential for resource reconfiguration and long-term growth (Teece, 2007; Gao et al., 2025). Innovation, as a signal independent of short-term financial performance, provides a new perspective for institutional investors to assess the quality of management and technological potential, which directly influences their stock holding decisions (Aghion et al., 2013). In this process, artificial intelligence can be used as an aid to R&D to unleash the innovation potential (Johnson et al., 2022; Roberts and Candi, 2024). Artificial intelligence can enhance innovation by increasing digital adaptability (Gao et al., 2025), improving the efficiency of firms’ operations (Rammer et al., 2022), and facilitating the efficiency of operations. Moreover, the enhancement of enterprise innovation ability releases the signal of long-term growth to institutional investors. Wang et al. (2023) indicated that the stronger the innovation ability of the company, the higher the proportion of institutional investor shareholding. Accordingly, we hypothesize as follows: H4. Artificial intelligence information disclosure will increase institutional investor ownership by improving the corporate innovation level. 4. Research design 4.1 Model specification To study the impact of corporate artificial intelligence information disclosure on the change in institutional investor ownership, we establish the following model: $$\:{dIO}_{i,t}={\beta\:}_{0}+{\beta\:}_{1}{dAI}_{i,t}+\sum\:{\beta\:}_{j}Control{s}_{i,t}+{\eta\:}_{i}+{\sigma\:}_{t}+{\epsilon\:}_{i,t}$$ 1 In Eq. ( 1 ), i represents the firm, t represents time, \(\:{dIO}_{i,t}\) is the change value of the dependent variable, institutional investor ownership, \(\:{dAI}_{i,t}\) is the change value of the core explanatory variable, artificial intelligence information disclosure index; \(\:Control{s}_{i,t}\) is a series of control variables; η i represents firm fixed effects, σ t represents year fixed effects, and \(\:{\epsilon\:}_{i,t}\) is the random error term. Additionally, we also cluster standard errors at the firm level. 4.2 Variable selection 4.2.1 Dependent variable The change value of institutional investor ownership (dIO): Following Bai et al.(2022), the institutional investor ownership constructed in this paper is the sum of the total shareholding ratio of institutional investors in listed companies, including the shareholding ratio of funds, QFIIs, brokers, insurance companies, social security funds, trust funds, finance companies and banks. 4.2.2 Independent variable The change value of the artificial intelligence information disclosure index (dAI): We construct the artificial intelligence information disclosure index based on the MD&A section in annual reports with the method of Zhang (2025) and Yao et al. (2024). First, we collected annual reports of listed companies from 2007 to 2023, and employed regular expressions to match keywords such as “Board of Directors Report” or “MD&A” at the beginning and “Important Matters” at the end to accurately extract the contents of the MD&A section in the annual report. Secondly, we manually selected 52 terms, including “Artificial Intelligence,” “Machine Learning,” “Internet of Things,” and others, as seed words. Next, we utilized the text from the annual report as the training corpus for the Word2vec technique and the Skip-gram model. Based on the cosine similarity between the seed words and the output words, we filtered the 10 words most semantically similar to each seed word. Subsequently, we removed duplicate words, words unrelated to artificial intelligence, and words with very low frequency. We ultimately obtained a list of 73 words as the artificial intelligence lexicon for this paper. Finally, “jieba” is used to segment the content of the MD&A. The artificial intelligence lexicon is added to the “jieba” segmentation module as a predefined proper noun. The frequency of occurrences of artificial intelligence keywords in the segmentation is then counted, and this count is used as the artificial intelligence information disclosure index. To further alleviate the endogeneity problem, this paper differentiates the dependent variable and the independent variable to eliminate the impact of fixed effects that do not change over time (Jiang et al., 2011; Panon, 2022; Thompson, 2025; Tie & Liu, 2025). Meanwhile, to facilitate the economic meaning of the coefficients, we standardized both the dependent variable and the independent variable, resulting in dIO and dAI (Babina et al., 2024). As can be seen from Fig. 1 , the development of artificial intelligence in China has shown a significant upward trend since 2007, and the fastest growth occurred in 2020. Maybe this acceleration is driven by the COVID-19 pandemic, which not only accelerates the use of artificial intelligence in healthcare (Annan and Qingge, 2025) but also forces enterprises to restructure their production mode as their employees adopt remote work to maintain business operations. Artificial intelligence has provided critical technological support for this transition (Aleem et al., 2023). 4.2.3 Control variables To avoid the problem of omitted variable bias, this paper refers to Li et al. (2016) and Zhang et al. (2023) to select control variables that may affect institutional investor ownership, ensuring an accurate assessment of the disclosure of artificial intelligence. Specifically, the following three types of variables are included: ( 1 ) corporate characteristics: firm size (Size), financial leverage (Lev), return on total assets (Roa), fixed assets ratio (Mortgage), market-to-book ratio (MB), and firm listing tenure (Age); ( 2 ) management characteristics: the percentage of independent directors (Independence), and shareholding ratio of the largest shareholder (Top1); ( 3 ) market performance aspects: annual turnover of outstanding shares (Turnover), and annual stock return (Return). The definitions of specific variables are shown in the appendix. 4.3 Sample source This paper takes 2007 as the starting year 3 and selects the data of China’s A-share listed companies in Shanghai and Shenzhen stock exchanges from 2007 to 2023 as the research sample. Then, we process the data as follows: ( 1 ) removing financial enterprises; ( 2 ) excluding samples of ST、PT and *ST enterprises; ( 3 ) dropping missing values; ( 4 ) winsorizing continuous variables at the 1% and 99% level; ( 5 ) deleting information transmission, software and information technology service industry and scientific research and technical service industry. We collect financial data from the China Stock Market and Accounting Research (CSMAR) database. The institutional investor ownership data originates from both the Wind database (Wind) and the CSMAR database. Additionally, we obtain stock turnover rates from the RESSET database. 4.4 Descriptive statistics analysis Table 1 presents the descriptive statistics for the primary variables. For Institutional investor ownership (IO), the minimum and maximum values are 0.000 and 0.665, respectively, indicating a significant gap between the extreme values. The mean and median values are 0.064 and 0.036, respectively, with a standard deviation of 0.077. This suggests that institutional investor ownership is generally low across most firms, with only a select few able to secure substantial institutional investment. Table 1 Descriptive statistics Variable N Mean SD Min Median Median IO 32256 0.064 0.077 0.000 0.036 0.665 AI 32256 4.353 11.804 0.000 1.000 430.000 Size 32256 22.365 1.296 19.858 22.182 26.220 Lev 32256 0.451 0.199 0.063 0.449 0.896 Roa 32256 0.038 0.061 -0.199 0.035 0.214 Mortgage 32256 0.232 0.163 0.002 0.202 0.719 MB 32256 0.636 0.253 0.128 0.632 1.185 Age 32256 2.362 0.641 0.693 2.485 3.367 Independence 32256 37.417 5.395 27.270 33.330 57.140 Top1 32256 0.346 0.148 0.089 0.324 0.749 Turnover 32256 5.872 0.900 3.362 5.966 7.722 Return 32256 0.175 0.615 -0.668 0.019 2.791 Regarding AI, the minimum and maximum values are 0 and 430, respectively. The mean and median values are 4.353 and 1, both significantly lower than the maximum value. This indicates that there are substantial disparities in corporate artificial intelligence information disclosure, with only a few companies inclined to disclose more information about artificial intelligence. All control variables exhibit data characteristics within optimal ranges, show no signs of outlier contamination, and are consistent with theoretical expectations. 4.5 Correlation analysis Table 2 presents the Pearson correlation coefficients for all variables, none of which exceed the threshold of 0.6, indicating that there is no evidence of severe multicollinearity among the explanatory variables. Table 2 Correlation analysis IO AI Size Lev Roa Mortgage MB Age Independence Top1 Turnover Return IO 1 AI 0.002 1 Size 0.189*** -0.029*** 1 Lev 0.004 -0.063*** 0.443*** 1 Roa 0.294*** -0.004 0.047*** -0.365*** 1 Mortgage -0.038*** -0.155*** 0.045*** 0.046*** -0.070*** 1 MB -0.193*** -0.076*** 0.558*** 0.370*** -0.221*** 0.096*** 1 Age -0.021*** -0.097*** 0.373*** 0.273*** -0.147*** 0.028*** 0.221*** 1 Independence -0.014** 0.045*** 0.022*** -0.013** -0.026*** -0.053*** -0.021*** -0.024*** 1 Top1 -0.084*** -0.084*** 0.210*** 0.046*** 0.134*** 0.076*** 0.134*** -0.048*** 0.035*** 1 Turnover -0.098*** -0.015*** -0.393*** -0.025*** -0.075*** 0.001 -0.260*** -0.199*** -0.039*** -0.160*** 1 Return 0.144*** -0.033*** -0.091*** 0.00200 0.140*** 0.035*** -0.340*** -0.072*** -0.029*** 0.013** 0.329*** 1 Note: This table reports the Pearson correlation coefficients between the main variables in the study. 5. Empirical results 5.1 Baseline regression results Table 3 presents the results of the baseline regression. To ensure the robustness of the regression results, column ( 1 ) shows that dAI is positively related to dIO at the 1% significance level, without any control variables included. Columns ( 2 ), ( 3 ), and ( 4 ) report the regression results with the control variables added sequentially in terms of firm characteristics, management characteristics, and market performance. The coefficients of dAI remain positive and statistically significant. Considering column ( 4 ) as the benchmark, the coefficient of dAI is 0.013 and is significantly positive at the 5% level, indicating that a one-standard-deviation increase in dAI corresponds to a 0.013-standard-deviation increase in dIO. This supports Hypothesis 1, which posits that increased disclosure of artificial intelligence will prompt institutional investors to increase their shareholdings. Table 3 Baseline regression results dIO ( 1 ) ( 2 ) ( 3 ) ( 4 ) dAI 0.015*** 0.014** 0.014** 0.013** (2.647) (2.483) (2.484) (2.449) Size -0.010 -0.010 -0.069*** (-0.778) (-0.761) (-5.598) Lev 0.064 0.065 -0.068 (1.061) (1.072) (-1.191) Roa 0.347*** 0.351*** -0.459*** (2.694) (2.722) (-3.631) Mortgage -0.162** -0.162** -0.251*** (-2.457) (-2.459) (-4.001) MB -0.836*** -0.835*** 0.135*** (-18.199) (-18.089) (2.956) Age -0.342*** -0.346*** -0.280*** (-11.611) (-11.427) (-9.894) Independence -0.001 -0.001 (-0.798) (-0.798) Top1 -0.044 -0.254*** (-0.551) (-3.305) Turnover -0.139*** (-12.920) Return 0.808*** (37.268) _cons -0.014*** 1.549*** 1.610*** 3.019*** (-349.254) (5.591) (5.686) (10.787) N 32256 32256 32256 32256 adj.R2 -0.054 -0.037 -0.038 0.058 Code FE Yes Yes Yes Yes Year FE Yes Yes Yes Yes Note: This table reports the results of benchmark regressions of change in artificial intelligence information disclosure (dAI) on change in institutional investor shareholdings (dIO). Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 5.2 Robustness analysis 5.2.1 Replacement of the independent variable Drawing on (Yao et al., 2024), we further analyzed the text of annual reports from listed companies, extracted the artificial intelligence word frequency based on the entire annual report data, constructed the indicator AI_words, and incorporated it into Eq. ( 1 ) as the replaced explanatory variable to re-estimate. The regression results are shown in column (1) of Table 4 , where the coefficient is 0.013, significant at the 5% level. This finding suggests that our results are robust to this alternative proxy for the artificial intelligence information disclosure index. Table 4 Robustness analysis ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) dIO dIO1 dIO2 dIO dIO dAI_words 0.013** (2.045) dAI 0.017*** 0.012* 0.013*** 0.013** (2.815) (1.679) (2.723) (2.242) Controls Yes Yes Yes Yes Yes Code FE Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes _cons 2.954*** 0.366 3.077*** 3.019*** 3.161*** (9.931) (1.005) (5.853) (7.852) (10.022) N 31352 31318 31318 32256 26304 adj.R2 0.057 0.044 0.035 0.058 0.060 Note: This table reports the regression results of the robustness tests. Column ( 1 ) replaces the independent variables. Columns ( 2 ) and ( 3 ) replace the dependent variables. Column ( 4 ) increases the clustering hierarchy to industry, and Column ( 5 ) removes municipality samples. In particular, dAI _words, dIO1, and dIO2 are variables obtained by differencing and standardizing the frequency of extracted artificial intelligence words based on annual reports, the ratio of the number of shares held by institutional investors to the total number of shares at the end of the year, and the ratio of the number of shares held by institutional investors to the number of shares outstanding at the end of the year. Code FE denotes firm fixed effects, and Year FE denotes year fixed effects. Standard errors are clustered at the firm level, and values in parentheses are t-values. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. 5.2.2 Replacement of the dependent variable Firstly, we follow the method of Li et al. (2016) and construct the ratio of the number of shares held by institutional investors to the total number of shares (IO1). Secondly, we follow the method of Zou et al. (2025) and construct the ratio of the number of shares held by institutional investors to the number of shares outstanding (IO2). Both variables are reintroduced as dependent variables in Eq. ( 1 ). The regression results are presented in columns ( 2 ) and ( 3 ) of Table 4 , and the coefficients for dAI remain positive and statistically significant (0.017 at 1% level and 0.012 at 10% level, respectively). These findings strengthen the support for Hypothesis 1, which posits a positive impact of artificial intelligence information disclosure. 5.2.3 Changing the clustering level We then shift the clustering level from the firm to the industry to control for factors that cannot be captured at the individual level. The subsequent regression results (column 4, Table 4 ) indicate that the coefficient of dAI is 0.013, which is significantly positive at the 1% level, suggesting that the conclusion remains robust after upgrading the clustering level. 5.2.4 Deletion of samples Considering that the special administrative status and city attributes of China’s four municipalities directly under the central government (Beijing, Tianjin, Shanghai, and Chongqing) might result in atypical estimates, we follow Tian et al. (2023) to rerun the model after excluding companies from those places in the robustness analysis. As shown in Column ( 5 ) of Table 4 , the coefficient for dAI is 0.013 and significant at the 5% level, confirming the robustness of our primary findings. 5.2.5 Quantile regression Recognizing that OLS regression can only capture the average impact while there is a large gap between firms, we adopt the quantile methodology pioneered by Koenker et al. (1978). We estimate at the 10%, 25%, 50%, 75%, and 90% quantiles controlling for firm and year-fixed effects (Agnese et al., 2024). As shown in Table 5 , the regression coefficients for dAI are 0.020, 0.013, 0.009, 0.010, and 0.029 for each quartile, respectively. All are significantly positive at the 5% or 1% level, which strongly supports the robustness of the baseline regression and suggests it is not influenced by extreme values. Furthermore, it can be observed that the coefficients exhibit a U-shaped pattern, with an initial decline followed by an increase. This indicates significant heterogeneity in the positive effect across institutional investor shareholdings. Table 5 Quantile regression ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) 10% 25% 50% 75% 90% dAI 0.020** 0.013** 0.009*** 0.010** 0.029*** (2.017) (2.518) (3.744) (2.097) (4.117) Controls Yes Yes Yes Yes Yes Code FE Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes N 32267 32267 32267 32267 32267 Note: This table reports the regression results of the robustness tests, which are quantile regressions with a change in estimation methodology. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, and values in parentheses are z-values. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. 5.3 Endogeneity tests 5.3.1 Instrumental variable approach and 2SLS estimation Although the baseline regression includes a series of variables that may affect institutional investors’ shareholding, there are still some unobservable factors related to the level of artificial intelligence that could lead to errors in the results of the benchmark regression. Therefore, we employ a two-stage least squares regression to mitigate the endogeneity issue. Firstly, following Zhang et al. (2024), we use the lagged period of the independent variable (L.dAI) as the instrumental variable. This instrument is correlated with the current period of the independent variable but is not correlated with the error term of the current period, thus satisfying both correlation and exogeneity. In addition, we also construct a Bartik instrumental variable regarding Goldsmith-Pinkham et al. (2020) and Acemoglu et al. (2022), which is the cross-multiplier of the difference between the firm’s artificial intelligence and the mean of the industry’s annual artificial intelligence with the exclusion of the firm’s data‌ and the lagged one-period term of the artificial intelligence, as shown in Eq. (2). where \(\:\stackrel{-}{A{I}_{i,t}}\) is the industry’s annual artificial intelligence mean after excluding firm i’s own data‌. This instrumental variable considers exogenous shocks at the industry level, the intensity of firms’ use of artificial intelligence technology within the industry, and the lagged level of artificial intelligence inputs to meet the criteria of relevance and exogeneity. Table 6 presents the regression results estimated using two-stage least squares (2SLS). Columns ( 1 ) and ( 3 ) display the regression outcomes for the two instrumental variables in the first stage, with F-values exceeding 10, satisfying the criterion for instrumental variable correlation. The Cragg-Donald Wald F-statistics are above the Stock-Yogo weak instrumental variable test’s 10% critical value of 16.38, passing the test for weak instrumental variables. The Kleibergen-Paap rk LM statistics are significant at the 1% level, rejecting the null hypothesis of “insufficient identification of instrumental variables,” thus indicating no under-identification issue. Columns ( 2 ) and ( 4 ) detail the second-stage regression results for the two instrumental variables, where the coefficients for dAI are significantly positive (0.046 and 0.065, respectively). This suggests that the baseline regression results remain robust even after addressing the endogeneity problem, thereby reaffirming that artificial intelligence information disclosure encourages institutional investors to increase their shareholdings. Table 6 Instrumental variable approach ( 1 ) ( 2 ) ( 3 ) ( 4 ) dAI dIO dAI dIO dAI 0.046* 0.065** (1.865) (2.429) L.dAI -0.248*** (-11.498) Bartik_iv 0.051*** (4.393) Controls Yes Yes Yes Yes Code FE Yes Yes Yes Yes Year FE Yes Yes Yes Yes N 27908 27908 27858 27858 adj.R2 0.109 0.108 F statistic 132.21*** 19.30*** Cragg-Donald Wald Fstatistic 1741.004*** 1694.386*** Kleibergen-Paap rk LM statistic 59.003*** 13.837*** Note: This table reports the results of two-stage least squares regressions, with columns ( 1 ) and ( 2 ) being the lagged one period for the first instrumental variable, dAI; and columns ( 3 ) and ( 4 ) being the second instrumental variable, the Barkit instrumental variable. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 5.3.2 The two-stage Heckman Firms’ decisions on whether to adopt artificial intelligence technology are typically not random and may be influenced by unobserved factors such as management preferences (Liu et al., 2025b) and technological capabilities. Companies that are more inclined to implement artificial intelligence may be more likely to disclose AI-related information in their annual reports, whereas those less focused on AI may provide less disclosure. This disparity could result in sample self-selection bias. Consequently, this paper employs the Heckman two-stage method to assess whether the sample exhibits a self-selection bias issue(Heckman, 1979). In the first stage, a dummy variable is constructed based on the mean value of artificial intelligence information disclosure for each industry-year, with values higher than the mean assigned to 1 and 0 otherwise (Chen et al., 2025). Additionally, the explanatory variable, lagged one period (L.dAI), is included as an exogenous variable. And we use a probit model to test whether the one-lagged control variables affect artificial intelligence (Jin et al., 2024). Next, we calculate the adjustment term, the inverse Mills ratio. In the second stage, the calculated inverse Mills ratio (IMR) is incorporated into the baseline regression Eq. ( 1 ) as a control variable to correct for estimation bias resulting from sample selection bias. As indicated in columns ( 1 ) and ( 2 ) of Table 7 , the impact of the IMR is negligible, and the coefficient for dAI remains significantly positive at the 1% level. This indicates that, after accounting for sample selection bias, the impact of artificial intelligence information disclosure on institutional investor ownership remains robust. Table 7 Endogeneity test ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) Phase I Phase II PSM Industry fixed effects Province fixed effects dAI 0.014*** 0.018* 0.013** 0.013** (2.653) (1.896) (2.398) (2.448) L.dAI -0.080*** (-8.399) imr 0.033 (0.278) Controls No Yes Yes Yes Yes Code FE No Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Ind FE No No No Yes No Province FE No No No No Yes _cons -1.052*** 2.596*** 2.964*** 3.254*** 3.019*** (-4.459) (7.237) (3.197) (11.222) (10.782) N 28158 27908 30964 32256 32256 adj.R2 / PseudoR2 0.025 0.061 0.334 0.057 0.057 Note: This table reports three endogeneity tests, columns ( 1 ) and ( 2 ) are Heckman two-stage; column ( 3 ) is PSM; column ( 4 ) adds industry fixed effects, and column ( 5 ) adds province fixed effects. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, Ind FE denotes firm fixed effects, Province FE denotes year fixed effects. and standard errors are clustered at the firm level. The values in the parentheses in the first column are z-values. The values in the other four bracketed columns are t-values. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. 5.3.3 Propensity score matching Referring to Li et al.(2023), we define the dummy variable (AI_dummy) by taking the mean value of artificial intelligence applications as the basis for division. Those applications with values higher than the mean are categorized into the treatment group (AI_dummy = 1), while those with values lower than the mean are categorized into the control group (AI_dummy = 0). We utilize Size, Lev, Roa, Age, Independence, Top1, Return, and R&D background of the executive team as matching variables for one-to-one nearest-neighbor propensity score matching with a caliper of 0.05. The balance test results indicate that the standardized deviations of the covariates post-matching are below 5%, and the t-test results support the initial hypothesis that the coefficients of the treatment and control groups are not significantly different. This suggests that the differences in characteristics between the experimental and control groups have been largely eliminated, and the matching effect is satisfactory. We rerun the regression using the matched samples, and the results are reported in column ( 3 ) of Table 7 . We find that the coefficient for dAI is 0.018, which is significantly positive at the 10% level, and we continue to observe a positive correlation between artificial intelligence information disclosure and institutional investor shareholding. 5.3.4 Industry-fixed effect and Province-fixed effect To address the issue of omitted variables, this paper incorporates industry fixed effects and province fixed effects into the regression analysis. The results are presented in columns ( 4 ) and ( 5 ) of Table 7 , indicating that the coefficients for dAI are significantly positive at the 5% level, corroborating the findings of the benchmark regression. 5.4 Heterogeneity test 5.4.1 Period heterogeneity The “Industry 4.0 policy” was initially proposed in Germany in 2011, with the goal of fostering the digital transformation of the manufacturing sector. Subsequently, the world has embarked on the fourth industrial revolution, and governments have expedited the advancement of artificial intelligence technology. Consequently, this paper uses 2011 as a pivotal year, dividing the sample into pre-2011 and post-2011 periods for heterogeneity analysis (Tao et al., 2024). As evident from columns ( 1 ) and ( 2 ) of Table 8 , the coefficients for dAI are insignificant prior to 2011 but significantly positive thereafter. This indicates that the impact of artificial intelligence on institutional investor ownership has become more pronounced post-2011. The potential reason for this shift could be that following 2011, China hastened the development of artificial intelligence, influenced by the fourth industrial revolution, to drive changes in the industrial structure. Table 8 Heterogeneity analysis ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) ( 7 ) ( 8 ) 2000–2010 2011–2020 Non-manufacturing Manufacturing High Competition Low Competition Low R&D background High R&D background dAI 0.135 0.012** -0.018 0.015*** 0.022** 0.003 0.001 0.019*** (1.396) (2.237) (-1.051) (2.723) (2.397) (0.395) (0.076) (3.260) Controls Yes Yes Yes Yes Yes Yes Yes Yes Code FE Yes Yes Yes Yes Yes Yes Yes Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes _cons 6.539** 2.525*** 3.083*** 3.425*** 3.039*** 3.534*** 3.197*** 3.349*** (2.319) (7.440) (5.995) (8.871) (6.713) (7.812) (7.450) (5.327) N 4007 28134 9827 22400 15573 16261 17467 13322 adj.R2 0.025 0.043 0.034 0.067 0.049 0.053 0.032 0.064 p-value 0.000*** 0.004*** 0.072* 0.060* Note: This table reports the results of the analysis of heterogeneity, with columns ( 1 ) and ( 2 ) based on the 2011 subgroups; columns ( 3 ) and ( 4 ) for the subgroup regressions of manufacturing firms and non-manufacturing firms; columns ( 5 ) and ( 6 ) for the subgroups based on median HHI; and columns ( 7 ) and ( 8 ) for the subgroups based on whether the chairman of the board of directors has an R&D background. Tests for differences between groups are Fisher’s combined tests, all with 500 samples. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 5.4.2 Industry heterogeneity Manufacturing serves as the cornerstone of the real economy and is a significant vehicle for the innovation strategy of artificial intelligence, playing a crucial role in the process of high-quality economic development. Given this, we categorize the sample into manufacturing and non-manufacturing industries (Wu et al., 2025). The results of the grouped regressions are presented in columns ( 3 ) and ( 4 ) of Table 8 , indicating that the impact of dAI on dIO is more pronounced in manufacturing firms. Conversely, no significant effect is noted in the non-manufacturing samples. The potential reason for this discrepancy is that manufacturing firms utilize AI technology to substitute for large equipment, thereby reducing fixed asset costs and enhancing productivity, which in turn encourages institutional investors to boost their shareholdings. In contrast, the non-manufacturing sector has a higher proportion of employees in sales and administrative roles, which involve tasks that are less conventional, such as social interactions-tasks that artificial intelligence currently cannot fully replicate (Wu et al., 2024). 5.4.3 Competitive heterogeneity Referring to Haushalter et al. (2007) and Ren et al. (2025), we measure the degree of competition in the industry using the Herfindahl index, or HHI, which is the cumulative sum of the squares of the ratios of each firm’s operating revenues to the industry’s total operating revenues. We stratify the sample into high and low competition groups based on the median of the HHI (Yang et al., 2025), where smaller values of the HHI indicate more intense competition. We observe that artificial intelligence information disclosure plays a more significant role in determining institutional investor ownership within firms operating in highly competitive industries (refer to column( 5 ) in Table 8 ). Conversely, there is no notable effect in the low competition group. The potential reason for this is that in highly competitive sectors, firms with increased artificial intelligence information disclosures tend to decrease information asymmetry, thereby enabling external institutional investors to access more information to inform their investment decisions. 5.4.4 Heterogeneity of board R&D background The board of directors is the key organization responsible for R&D expenditure decisions. Directors with specific technology backgrounds can provide unique insights to guide corporate R&D investments. They communicate and collaborate with the CEO to ensure that the company’s R&D strategy aligns with the overall corporate strategy and play a crucial role in promoting the development of corporate artificial intelligence. Consequently, we follow Li et al. (2025) to calculate the ratio of directors with R&D backgrounds, grouping the observations according to the median of this indicator. As evident from columns ( 7 ) and ( 8 ) of Table 9 , the coefficient of dAI on dIO is 0.019 and significant at the 1% level in the sample with a high R&D background. However, it is not significant in the sample with a low R&D background. This suggests that the impact of artificial intelligence information disclosure on institutional investor shareholdings is greater in the sample with a high R&D background. A possible reason for this is that directors with a technological background are more inclined to allocate more resources to strengthen their firm’s technological capabilities (Ginesti et al., 2025) and to promote the development of artificial intelligence technology. Table 9 Institutional investor site visits ( 1 ) ( 2 ) ( 3 ) ( 4 ) dSV1 dSV2 dSV3 SV4 dAI 0.022*** 0.017** 0.016** 0.033*** (3.294) (2.391) (2.310) (5.464) Controls Yes Yes Yes Yes Code FE Yes Yes Yes No Year FE Yes Yes Yes No _cons 0.749** 0.294 0.239 0.233 (2.022) (0.760) (0.626) (0.562) N 25412 25412 25412 26925 adj.R2 /Pseudo R2 -0.011 -0.028 -0.032 0.0682 Note: This table reports the results of the mediation mechanism for institutional investor fieldwork. dSV1, dSV2, and dSV3 are differenced and standardized variables for the number of institutional surveys, the number of institutional surveys, and the number of people researched. SV4 is a dummy variable that equals 1 if an institutional investor site visit occurs to the enterprise in the current year, and 0 otherwise. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, and standard errors are clustered at the firm level. The values in parentheses in columns ( 1 ), ( 2 ), and ( 3 ) are t-values, and the values in parentheses in column ( 4 ) are z-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 5.5 Mechanism analysis The previous theoretical analysis indicates that the disclosure of artificial intelligence information can encourage institutional investor shareholdings by prompting institutional investors to conduct fieldwork, enhancing information transparency, and fostering innovation. Consequently, this paper concentrates on the impact of artificial intelligence information disclosure on these three avenues and constructs the following model to test the mechanism of action: $$\:{dM}_{i,t}={\varphi\:}_{0}+{\varphi\:}_{1}{dAI}_{i,t}+{\varphi\:}_{2}Control{s}_{i,t}+{\lambda\:}_{i}+{\mu\:}_{t}+{\epsilon\:}_{i,t}$$ 3 where subscripts i and t denote enterprises and years, respectively; M denotes mechanism variables; and other variable settings are consistent with the benchmark regression Eq. ( 1 ). 5.5.1 Institutional investor site visits We follow the methodologies of Jiang et al. (2018) and Cao et al. (2024) to select the number of institutional investor visits (SV1) as the first proxy variable for fieldwork; refer to Zhao et al. (2023) to select the number of institutions visited (SV2) as the second proxy variable for fieldwork; following Xu et al. (2025), the number of institutional investors’ research (SV3) is selected as the third proxy variable for fieldwork; regarding (Lai et al., 2022), the dummy variable for institutional investors’ fieldwork (SV4) is selected as the fourth proxy variable for fieldwork, and takes the value of 1 if the company has institutional investors’ fieldwork in that year, and 0 otherwise. As indicated in Table 9 , corporate artificial intelligence information disclosure influences the indicators across various dimensions of institutional investor fieldwork. Initially, the disclosure of artificial intelligence information notably boosts the quantity of institutional research. This suggests that following enterprises’ release of artificial intelligence data, institutional investors are more inclined to communicate and interact with these companies, thereby gaining a better understanding of their business operations and the development of artificial intelligence technologies. Secondly, the number of institutional research studies sees a significant rise, implying that the disclosure of artificial intelligence attracts a broader range of institutional investors with diverse backgrounds and investment strategies to conduct their own investigations. Thirdly, there is a notable increase in the number of researchers, signifying that institutions are eager to allocate more personnel to participate in the research, which also substantially enhances the likelihood of field research taking place. In summary, artificial intelligence information disclosure fosters various aspects of institutional investors’ fieldwork activities. Following site visits, institutional investors will then determine whether to augment or reduce their stock holdings. 5.5.2 Information transparency Firstly, concerning Clarkson et al. (2008), we construct the Janis-Fadner coefficient, an indicator of media attention, as the first proxy variable of information transparency, as shown in Eq. ( 4 ): $$\:J-F=\left\{\begin{array}{c}\frac{{e}^{2}-ec}{{t}^{2}}\:\:\:\:\:\:\:\:\:\:\:\:\:\:if\:e>c\\\:\frac{{ec-c}^{2}}{{t}^{2}}\:\:\:\:\:\:\:\:\:\:\:\:\:\:if\:e<c\\\:0\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:if\:e=c\end{array}\right.$$ 4 where e is the number of positive media reports, c is the number of negative media reports, and t is the sum of the number of positive reports and the number of negative reports. The value of the J-F coefficient ranges from − 1 to 1. The more positive reports about the enterprise, the closer the J-F coefficient is to 1, and the less pressure the enterprise faces from media monitoring. When there are more negative reports about the enterprise, the J-F coefficient is close to -1, indicating that the enterprise faces more media monitoring pressure. Secondly, following Weng et al. (2024), we select the number of analyst coverage, the number of analysts tracking a company in a year, as the second proxy variable of information transparency. The regression results in columns ( 1 ) and ( 2 ) of Table 10 indicate that firm-level artificial intelligence information disclosure significantly enhances media monitoring and analyst coverage. Additionally, the results suggest that artificial intelligence information disclosure fosters institutional investor ownership by enhancing information transparency. Table 10 Information Transparency ( 1 ) ( 2 ) dmedia danalyst dAI 0.015** 0.022** (2.094) (2.542) Controls Yes Yes Code FE Yes Yes Year FE Yes Yes _cons 0.618** 1.647*** (2.401) (3.364) N 29967 19012 adj.R2 -0.070 0.092 Note: This table reports the results of the mediation mechanism of information transparency. dmedia and danalyst are differenced and standardized variables for the J-F coefficient and the number of analyst coverage. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and the values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. The possible reasons are that the media enables institutional investors to make objective judgments about the development of corporate artificial intelligence through reports on AI investments and AI patents. With their extensive knowledge, analysts can provide institutional investors with professional reports, which are crucial for them to decide whether to increase their stock holdings (Li et al., 2024). 5.5.3 Innovation In this paper, we reference existing research (Yuan et al., 2018) and calculate the sum of invention patents, utility model patents, and design patents to measure the innovation level of enterprises (patents). Table 11 presents the results of the mediation test for the innovation level of enterprises. We report that dAI is significant at the 10% level, indicating that changes in artificial intelligence information disclosure can promote the innovation level of enterprises. Innovation, as a key indicator of the long-term development of enterprises, has an important impact on the stock selection decision of institutional investors. When an enterprise’s innovation ability is enhanced, it not only significantly increases the success rate of enterprise R&D but also sends a positive signal to institutional investors about the long-term viability of the enterprise, which encourages them to increase their stock holdings. Table 11 Innovation dpatent dAI 0.031* (1.658) Controls Yes Code FE Yes Year FE Yes _cons 0.252 (1.461) N 31352 adj. R2 -0.102 Note: This table reports the results of the innovation mediation mechanism. dpatent is the difference-in-differences and standardization after summing invention patents, utility model patents, and design patents (patent). Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and the values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 6. Further analysis Although institutional investors share certain commonalities, they must not be viewed as a homogeneous group (Yan et al., 2009). Though institutional investors emphasize the long-term value of firms, different types of institutional investors have varying levels of concern for the long-term value of firms (Cox et al.,2011). To investigate the impact of artificial intelligence information disclosure on institutional investors, this paper categorizes them into short-term and long-term institutional investors, as well as stress-sensitive and stress-resistant institutional investors. 6.1. Short- and long-term institutional investors Referring to Yan et al.(2009), we construct the indicators of short-term institutional investors and long-term institutional investors according to the following steps. Firstly, we calculate the cumulative total stock assets bought and sold by each institutional investor in each half-year from 2007 to 2023. For a given institutional investor k, the cumulative total assets of stocks bought and sold per half-year are computed as follows, respectively: $$\:\begin{array}{c}C{R}_{-}bu{y}_{k,t}=\sum\:_{i=1}^{N}\:\left|{S}_{k,i,t}{P}_{i,t}-{S}_{k,i,t-1}{P}_{i,t-1}-{S}_{k,i,t-1}{\Delta\:}{P}_{i,t}\left|\right({S}_{k,i,t}⩾{S}_{k,i,t-1})\right.\end{array}$$ 5 $$\:\begin{array}{c}C{R}_{-}{sell}_{k,t}=\sum\:_{i=1}^{N}\:\left|{S}_{k,i,t}{P}_{i,t}-{S}_{k,i,t-1}{P}_{i,t-1}-{S}_{k,i,t-1}{\Delta\:}{P}_{i,t}\left|\right({S}_{k,i,t}<{S}_{k,i,t-1})\right.\end{array}$$ 6 \(\:{P}_{i,t}\) and \(\:{P}_{i,t-1}\) are the stock prices of the listed company \(\:i\) in period t and t-1, and \(\:{S}_{k,i,t}\) and \(\:{S}_{k,i,t-1}\) denote the number of shares held by institutional investor k in listed company \(\:i\) in period t and t-1, respectively. \(\:C{R}_{-}bu{y}_{k,t}\) and \(\:C{R}_{-}{sell}_{k,t}\) denote the cumulative market value of shares bought and sold by institutional investor k in period t, respectively. Secondly, the turnover rate of institutional investor k in period t is defined as: $$\:C{R}_{k,t}=\frac{2min(C{R}_{-}bu{y}_{k,t},C{R}_{-}sel{l}_{k,t})}{\sum\:_{i=1}^{{N}_{k}}\:({S}_{k,i,t}{P}_{i,t}+{S}_{k,i,t}{P}_{i,t-1})}$$ 7 Thirdly, the average turnover rate of institutional investor k is calculated, whose average turnover rate is defined as the average turnover rate over the last two years (4 semiannual periods): $$\:AV{G}_{-}C{R}_{k,t}=\frac{1}{4}\sum\:_{j=0}^{3}\:C{R}_{k,t-j}$$ 8 Fourthly, institutional investors are categorized based on their average turnover rate. To avoid the influence of outliers, those holding investments for two or more years between 2007 and 2023 are divided into three groups following a sorting by their average turnover rate. The group exhibiting the highest average turnover rate is classified as short-term institutional investors, while the group with the lowest turnover rate is classified as long-term institutional investors. Accordingly, the percentage of ownership by short-term institutional investors (SIO) and the percentage of ownership by long-term institutional investors (LIO) in each stock are calculated separately. From columns ( 1 ) and ( 2 ) of Table 12 , the coefficient of dAI is significantly positive for the regression result of dLIO, but insignificant for dSIO. This suggests that long-term institutional investors are more attentive to the long-term value investment of companies and focus on the long-term value that artificial intelligence technology brings (Cremers et al., 2020). Table 12 Further analysis ( 1 ) ( 2 ) ( 3 ) ( 4 ) dSIO dLIO dNIS dIS dAI 0.007 0.010* -0.004 0.010** (1.207) (1.855) (-0.879) (2.068) Controls Yes Yes Yes Yes Code FE Yes Yes Yes Yes Year FE Yes Yes Yes Yes _cons 3.678*** -0.041 0.458 3.524*** (11.751) (-0.145) (1.501) (11.447) N 31235 31235 31235 31235 adj.R2 0.042 -0.063 -0.081 0.109 Note: This table reports the results of the further analysis. The short-term institutional investors ownership (SIO) is first differenced and then standardized to obtain dSIO. The long-term institutional investors ownership (LIO) is first differenced and then standardized to obtain dLIO. The pressure-sensitive institutional investors ownership (NIS) is first differenced and then standardized to obtain dNIS. The pressure-resistant institutional investors ownership (IS) is first differenced and then standardized to obtain dIS. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and the values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. 6.2 Stress-sensitive, stress-resistant institutional investors Furthermore, following Brickley et al. (1988) and Liu et al. (2024), institutional investors are classified into pressure-sensitive (NIS) and pressure-resistant (IS) categories. Pressure-sensitive institutional investor ownership encompasses the combined shareholding percentages of insurance companies, trust companies, comprehensive brokerage firms, and corporate annuities that engage directly with the company. Their investment decisions are susceptible to management influence, and their inclination to participate in corporate governance is typically low. In contrast, pressure-resistant institutional investor ownership includes the combined shareholding percentages of Securities Investment funds, social security funds, and QFII. These investors have no direct business relations with the company and often base their voting decisions on independent analysis, which tends to enhance the corporate governance standards. This category of institutional investor is primarily focused on the long-term value of the investment. The results in columns ( 3 ) and ( 4 ) of Table 12 suggest that artificial intelligence information disclosure attracts stress-resistant institutional investors, rather than stress-sensitive ones, to increase their stock holdings. On the one hand, artificial intelligence information disclosure can reduce the degree of information asymmetry, providing stress-resistant institutional investors with more information about artificial intelligence. On the other hand, it offers signals about the long-term development of firms, indicating that these firms possess unique resource advantages and robust R&D capabilities. This attracts investors to increase their stock holdings, allowing them to benefit from the future development of artificial intelligence. 7. Conclusion As a pivotal force in the new wave of technological revolution and industrial transformation, artificial intelligence significantly contributes to fostering corporate innovation, enhancing productivity, and propelling economic growth. However, the impact of artificial intelligence information disclosure on institutional investor shareholding remains inconclusive. This paper examines how changes in corporate artificial intelligence information disclosure influence institutional investor shareholdings. Based on the dataset of Chinese listed companies from 2007 to 2023, we find that the disclosure of artificial intelligence information by corporations encourages institutional investors to augment their equity stakes. This conclusion is supported by five robustness tests and a battery of endogeneity assessments. The robustness tests encompassed substituting dependent variables, substituting independent variables, elevating the clustering level to the industry level, excluding municipal samples, and altering the regression approach. To mitigate endogeneity issues, we employed instrumental variables in a two-stage least squares regression. To counteract sample self-selection bias, we conducted a Heckman two-stage test. Additionally, we utilized Propensity Score Matching (PSM) to address selection bias in the sample. We also considered the potential problem of omitted variables by incorporating industry and province fixed effects, respectively. Next, we perform heterogeneity analyses, examining time heterogeneity, industry heterogeneity, competition heterogeneity, and director technology background heterogeneity. We find that the contribution of changes in artificial intelligence information disclosure to changes in institutional investor ownership is more significant in post-2011 samples, within the manufacturing industry, in environments with a high degree of competition, and among companies with a high proportion of directors with technology backgrounds. Following this, we conduct the mechanism analysis to investigate whether artificial intelligence information disclosure affects institutional investor ownership by promoting site visits from institutional investors, enhancing information transparency, and improving innovation. This serves to test the three proposed hypotheses. Finally, we further analyze the impact of artificial intelligence information disclosure on the shareholding ratios of different types of institutional investors. Specifically, we classify institutional investors into long-term and short-term investors, as well as stress-resistant and stress-sensitive investors. As evidenced by the literature, the decision of institutional investors to hold shares in companies for investment purposes is, to a certain extent, dependent on the development prospects, operational capabilities, and financial status of those enterprises. Furthermore, their emphasis on advanced technology during the development process is underscored by the disclosure of artificial intelligence. This paper, therefore, contributes to the existing literature on institutional investors’ stockholding and addresses the gap concerning whether artificial intelligence influences institutional stockholding. Declarations Funding: This work was partially supported by the National Social Science Fund of China [No. 20BJY260], and BTBU Digital Business Platform Project by BMEC. Author Contribution Feng Zhao: Conceptualization, Writing - Review & Editing, and Funding acquisition;Dandan Nan: Data curation, Methodology, Software, and Writing;Yao Wu: Methodology, Validation, Writing - Review & Editing and Supervision Data Availability Data are available upon request by email. References Acemoglu D, Autor D, Hazell J, et al., 2022. Artificial intelligence and jobs: evidence from online vacancies[J]. Journal of Labor Economics, 40: S293-S340. Aggarwal R, Erel I, Ferreira M, et al., 2011. Does governance travel around the world? Evidence from institutional investors[J]. Journal of Financial Economics, 100(1): 154-181. Aghion P, Van Reenen J, Zingales L, 2013. Innovation and institutional ownership[J]. American Economic Review, 103(1): 277-304. Agnese P, Arduino F R, Bruno E, et al., 2024. On the road to sustainability: the role of board characteristics in driving ESG performance in Africa[J]. Socio-Economic Planning Sciences, 95: 101994. Ahmed I, Jeon G, Piccialli F, 2022. From artificial intelligence to explainable artificial intelligence in industry 4.0: a survey on what, how, and where[J]. IEEE Transactions on Industrial Informatics, 18(8): 5031-5042. Aleem M, Sufyan M, Ameer I, et al., 2023. Remote work and the COVID-19 pandemic: an artificial intelligence-based topic modeling and a future agenda[J]. Journal of Business Research, 154: 113303. Alonso C, Berg A, Kothari S, et al., 2022. Will the AI revolution cause a great divergence?[J]. Journal of Monetary Economics, 127: 18-37. Annan R, Qingge L, 2025. Artificial intelligence in COVID-19 research: a comprehensive survey of innovations, challenges, and future directions[J]. Computer Science Review, 57: 100751. Appel I R, Gormley T A, Keim D B, 2016. Passive investors, not passive owners[J]. Journal of Financial Economics, 121(1): 111-141. Armstrong C S, Core J E, Taylor D J, et al., 2011. When does information asymmetry affect the cost of capital?[J]. Journal of Accounting Research, 49(1): 1-40. Babina T, Fedyk A, He A, et al., 2024. Artificial intelligence, firm growth, and product innovation[J]. Journal of Financial Economics, 151: 103745. Bag S, Pretorius J H C, Gupta S, et al., 2021. Role of institutional pressures and resources in the adoption of big data analytics powered artificial intelligence, sustainable manufacturing practices and circular economy capabilities[J]. Technological Forecasting and Social Change, 163: 120420. Bahoo S, Cucculelli M, Qamar D, 2023. Artificial intelligence and corporate innovation: a review and research agenda[J]. Technological Forecasting and Social Change, 188: 122264. Bai X, Han J, Ma Y, et al., 2022. ESG performance, institutional investors’ preference and financing constraints: empirical evidence from China[J]. Borsa Istanbul Review, 22: S157-S168. Balakrishnan K, Billings M B, Kelly B, et al., 2014. Shaping liquidity: on the causal effects of voluntary disclosure[J]. The Journal of Finance, 69(5): 2237-2278. Barney J, 1991. Firm resources and sustained competitive advantage[J]. Journal of Management, 17(1): 99-120. Begenau J, Farboodi M, Veldkamp L, 2018. Big data in finance and the growth of large firms[J]. Journal of Monetary Economics, 97: 71-87. Belev B, 2003. Institutional investors in bulgarian corporate governance reform: obstacles or facilitators?[J]. Journal of World Business, 38(4): 361-374. Belhadi A, Kamble ,Sachin, Fosso Wamba ,Samuel, et al., 2022. Building supply-chain resilience: an artificial intelligence-based technique and decision-making framework[J]. International Journal of Production Research, 60(14): 4487-4507. Berente N, Gu B, Recker J, et al., 2021. Managing artificial intelligence[J]. MIS Quarterly. Boot A, Hoffmann P, Laeven L, et al., 2021. Fintech: what’s old, what’s new?[J]. Journal of Financial Stability, 53: 100836. Brickley J A, Lease R C, Smith C W, 1988. Ownership structure and voting on antitakeover amendments[J]. Journal of Financial Economics, 20: 267-291. Bushee B J, Core J E, Guay W, et al., 2010. The role of the business press as an information intermediary[J]. Journal of Accounting Research, 48(1): 1-19. Cao J, Dong D, Yue S, 2024. Institutional investors’ site visits and firms’ financial distress[J]. Research in International Business and Finance, 67: 102150. Chang Z, 2025. Influence of artificial intelligence on the financial labor market[J]. Finance Research Letters, 80: 107039. Chen K, Chen X, Wang Z ao, et al., 2024. Does artificial intelligence promote common prosperity within enterprises? —evidence from chinese-listed companies in the service industry[J]. Technological Forecasting and Social Change, 200: 123180. Chen Y, Du L, Zhang B, et al., 2025. The impact of artificial intelligence on the sustainability of international trade enterprises[J]. International Review of Economics & Finance, 101: 104136. Cheng H, Huang D, Luo Y, 2020. Corporate disclosure quality and institutional investors’ holdings during market downturns[J]. Journal of Corporate Finance, 60: 101523. Chishti M Z, Dogan E, Binsaeed R H, 2024. Can artificial intelligence and green finance affect economic cycles?[J]. Technological Forecasting and Social Change, 209: 123740. Chopra R, Sharma G D, Pereira V, 2024. Identifying bulls and bears? A bibliometric review of applying artificial intelligence innovations for stock market prediction[J]. Technovation, 135: 103067. Choudhury T, 2025. US sectors and geopolitical risk: the investor’s perspective[J]. Finance Research Letters, 73: 106690. Chung H, Judge W Q, Li Y H, 2015. Voluntary disclosure, excess executive compensation, and firm value[J]. Journal of Corporate Finance. Chung K H, Zhang H, 2011. Corporate governance and institutional ownership[J]. Journal of Financial and Quantitative Analysis, 46(1): 247-273. Clarkson P M, Li Y, Richardson G D, et al., 2008. Revisiting the relation between environmental performance and environmental disclosure: an empirical analysis[J]. Accounting, Organizations and Society, 33(4-5): 303-327. Cohen L, Malloy C, Nguyen Q, 2020. Lazy prices[J]. The Journal of Finance, 75(3): 1371-1415. Cox P, Wicks P G, 2011. Institutional interest in corporate responsibility: portfolio evidence and ethical explanation[J]. Journal of Business Ethics. Cremers M, Pareek A, Sautner Z, 2020. Short-term investors, long-term investments, and firm value: evidence from russell 2000 index inclusions[J]. Management Science, 66(10): 4535-4551. Culot G, Podrecca M, Nassimbeni G, 2024. Artificial intelligence in supply chain management: a systematic literature review of empirical studies and research directions[J]. Computers in Industry, 162: 104132. Czarnitzki D, Fernández G P, Rammer C, 2023. Artificial intelligence and firm-level productivity[J]. Journal of Economic Behavior & Organization, 211: 188-205. Das K Ch, Rani N, Bodhi R, et al., 2025. Do robots impact artificial intelligence (AI)-related employment? Evidence from a cross-national study[J]. International Journal of Manpower. Ding M, Gao Q, 2025. The impact of artificial intelligence technology application on total factor productivity in agricultural enterprises: evidence from China[J]. Economic Analysis and Policy, 86: 399-415. Du K, Zhao Y, Mao R, et al., 2025. Natural language processing in finance: a survey[J]. Information Fusion, 115: 102755. Duan D, Chen S, Feng Z, et al., 2023. Industrial robots and firm productivity[J]. Structural Change and Economic Dynamics, 67: 388-406. Dyck A, Lins K V, Roth L, et al., 2019. Do institutional investors drive corporate social responsibility? International evidence[J]. Journal of Financial Economics, 131(3): 693-714. Fan X, Wu Y, Zhou Y, et al., 2025. How does artificial intelligence shock affect labor income distribution? Evidence from China[J]. Pacific-Basin Finance Journal, 90: 102691. Fiorillo P, Meles A, Salerno D, et al., 2024. Geopolitical turmoil and investor green preference: evidence from the corporate bond market[J]. Journal of International Money and Finance, 149: 103218. Flammer C, Toffel M W, Viswanathan K, 2021. Shareholder activism and firms’ voluntary disclosure of climate change risks[J]. Strategic Management Journal, 42(10): 1850-1879. Francis B B, Hasan I, Zhu Y, 2021. The impact of political uncertainty on institutional ownership[J]. Journal of Financial Stability, 57: 100921. Fu Y, Shen Y, Song M, et al., 2024. Does artificial intelligence reduce corporate energy consumption? New evidence from China[J]. Economic Analysis and Policy, 83: 548-561. Gao L, Wang J, 2025. Can artificial intelligence reduce energy vulnerability? Evidence from an international perspective[J]. Energy Economics, 145: 108491. Gao Y, Liu S, Yang L, 2025. Artificial intelligence and innovation capability: a dynamic capabilities perspective[J]. International Review of Economics & Finance, 98: 103923. Gaver J J, Gaver K M, 1993. Additional evidence on the association between the investment opportunity set and corporate financing, dividend, and compensation policies[J]. Journal of Accounting and Economics, 16(1-3): 125-160. Ginesti G, Spanò R, Campa D, et al., 2025. Board composition, audit committee characteristics, and R&D expenditure[J]. European Management Journal, 43(1): 115-129. Giordino D, Sestino A, Zàmek D, et al., 2025. The impact of institutional ownership structure on corporate ESG performance: empirical evidence for the european capital market[J]. Review of Managerial Science. Goh L, Liu X, Tsang A, 2020. Voluntary disclosure of corporate political spending[J]. Journal of Corporate Finance, 61: 101403. Goldfarb A, Tucker C, 2019. Digital economics[J]. Journal of Economic Literature, 57(1): 3-43. Goldsmith-Pinkham P, Sorkin I, Swift H, 2020. Bartik instruments: what, when. Why, and how[J]. American Economic Review, 110(8): 2586-2624. Deb G S, 2018. Institutional investors and firm characteristics: new evidence from India[J]. Research in International Business and Finance, 46: 30-42. Haefner N, Wincent J, Parida V, et al., 2021. Artificial intelligence and innovation management: a review, framework, and research agenda✰[J]. Technological Forecasting and Social Change, 162: 120392. Han Q, Li C, Jin Y, 2025. The impact of intellectual property protection on the development of artificial intelligence in enterprises[J]. International Review of Financial Analysis: 104179. Hao (Grace) Qing, Li A, 2024. Economic policy uncertainty and institutional portfolio investment[J]. Financial Management, n/a(n/a). Haushalter D, Klasa S, Maxwell W, 2007. The influence of product market dynamics on a firm’s cash holdings and hedging behavior☆[J]. Journal of Financial Economics, 84(3): 797-825. He G, Li Z, Yu L, et al., 2025. Does commercial reform embracing digital technologies mitigate stock price crash risk?[J]. Journal of Corporate Finance, 91: 102741. Heckman J J, 1979. Sample selection bias as a specification error[J]. Econometrica, 47(1): 153-161. Hong H, Kubik J D, Stein J C, 2008. The only game in town: stock-price consequences of local bias$[J]. Journal of Financial Economics. Huang J, Han F, Li Y, 2025. The impact of institutional investors’ ESG concerns on corporate ESG disclosure: evidence from site visits[J]. Finance Research Letters, 76: 106957. Huang M, Li Q, Li B, 2025. From algorithms to green growth: can artificial intelligence drive enterprise energy transformation?[J]. Economic Analysis and Policy, 85: 1846-1866. Huang Q, Li Y, Lin M, et al., 2022. Natural disasters, risk salience, and corporate ESG disclosure[J]. Journal of Corporate Finance, 72: 102152. Iacucci M, Santacroce G, Yasuharu M, et al., 2025. Artificial intelligence–driven personalized medicine: transforming clinical practice in inflammatory bowel disease[J]. Gastroenterology: S0016508525004949. Jia H, Che W, 2025. Institutional investor stability, executive equity incentives, and corporate innovation[J]. Finance Research Letters, 83: 107691. Jiang C X, Kim J C, Zhou D, 2011. Liquidity, analysts, and institutional ownership[J]. International Review of Financial Analysis, 20(5): 335-344. Jiang X, Yuan Q, 2018. Institutional investors’ corporate site visits and corporate innovation[J]. Journal of Corporate Finance, 48: 148-168. Jin C, Monfort A, Chen F, et al., 2024. Institutional investor ESG activism and corporate green innovation against climate change: exploring differences between digital and non-digital firms[J]. Technological Forecasting and Social Change, 200: 123129. Johnson P C, Laurell C, Ots M, et al., 2022. Digital innovation and the effects of artificial intelligence on firms’ research and development – automation or augmentation, exploration or exploitation?[J]. Technological Forecasting and Social Change, 179: 121636. Joseph J, Wilson A J, 2018. The growth of the firm: an attention-based view[J]. Strategic Management Journal, 39(6): 1779-1800. Kacperczyk M T, Nosal J, Wang T, 2022. Global volatility and firm-level capital flows[J]. SSRN Electronic Journal. Koenker R, Bassett G, 1978. Regression quantiles[J]. Econometrica, 46(1): 33-50. Kong D, Lin C, Liu S, et al., 2021. Whose money is smart? Individual and institutional investors’ trades based on analyst recommendations[J]. Journal of Empirical Finance, 62: 234-251. Krueger P, Sautner Z, Starks L T, 2019. The importance of climate risks for institutional investors[J]. Lai S, Li X, Liu S, et al., 2022. Institutional investors’ site visits and corporate employment decision-making[J]. Journal of Contemporary Accounting & Economics, 18(3): 100332. Laksmana I, Harjoto M A, Kim H, 2023. Managing disclosure of political risk: the case of socially responsible firms[J]. Journal of Business Research, 154: 113366. Lee C C, Qin S, Li Y, 2022. Does industrial robot application promote green technology innovation in the manufacturing industry?[J]. Technological Forecasting and Social Change, 183: 121893. Lewellen J, Lewellen K, 2022. Institutional investors and corporate governance: the incentive to Be engaged[J]. The Journal of Finance, 77(1): 213-264. Li C, Xu Y, Zheng H, et al., 2023. Artificial intelligence, resource reallocation, and corporate innovation efficiency: evidence from China’s listed companies[J]. Resources Policy, 81: 103324. Li G, Ren K, Qiao Y, et al., 2024. From framing to priming: how does media coverage promote ESG preferences of institutional investors[J]. Emerging Markets Review, 63: 101220. Li G, Shao Y, 2023. How do top management team characteristics affect digital orientation? Exploring the internal driving forces of firm digitalization[J]. Technology in Society, 74: 102293. Li H, Lu Z, Zhang Z, et al., 2025. How does artificial intelligence affect manufacturing firms’ energy intensity?[J]. Energy Economics, 141: 108109. Li H, Wu H, Rao J, 2025. Impact of artificial intelligence on corporate green transformation[J]. Finance Research Letters, 80: 107427. Li J, Li M, Wang X, et al., 2021. Strategic directions for AI: the role of cios and boards of directors[J]. MIS Quarterly, 45(3): 1603-1643. Li J, Ma S, Qu Y, et al., 2023. The impact of artificial intelligence on firms’ energy and resource efficiency: empirical evidence from China[J]. Resources Policy, 82: 103507. Li J, Wu T, Hu B, et al., 2025. Artificial intelligence and corporate ESG performance[J]. International Review of Financial Analysis, 102: 104036. Li W, Lu X, 2016. Institutional interest, ownership type, and environmental capital expenditures: evidence from the most polluting chinese listed firms[J]. Journal of Business Ethics, 138(3): 459-476. Li Y, Goodell J W, Shen D, 2023. Market reaction to climate risk report disclosures: the roles of investor attention and sentiment[J]. Finance Research Letters, 58: 104290. Li Y, Zhang T, Tao Y, et al., 2025. Technology-based boards and bank operational efficiency: Mediating effects of risk-taking[J]. Finance Research Letters, 76: 106924. Lin F, Wu C M, Fang T Y, et al., 2014. The relations among accounting conservatism, institutional investors and earnings manipulation[J]. Economic Modelling, 37: 164-174. Lin Y C, Padliansyah R, Lu Y H, et al., 2025. Bankruptcy prediction: integration of convolutional neural networks and explainable artificial intelligence techniques[J]. International Journal of Accounting Information Systems, 56: 100744. Liu J, Xiong X, Gao Y, et al., 2023. The impact of institutional investors on ESG: evidence from China[J]. Accounting & Finance, 63(S2): 2801-2826. Liu R, Luo W, Su R, 2025. How does cultural diversity influence corporate AI development?[J]. Finance Research Letters, 81: 107506. Liu Y, Jin X M, Ly K C, et al., 2024. The relationship between heterogeneous institutional investors’ shareholdings and corporate ESG performance: evidence from China[J]. Research in International Business and Finance, 71: 102457. Liu Y, Shen F, Guo J, et al., 2025. Can artificial intelligence technology improve companies’ capacity for green innovation? Evidence from listed companies in China[J]. Energy Economics, 143: 108280. Liu Y, Wang B, Song Z, 2025. Promoting or inhibiting: the impact of artificial intelligence application on corporate environmental performance[J]. International Review of Financial Analysis, 97: 103872. Luqman A, Zhang Q, Talwar S, et al., 2024. Artificial intelligence and corporate carbon neutrality: a qualitative exploration[J]. Business Strategy and the Environment, 33(5): 3986-4003. Lyssimachou D, Bilinski P, 2023. Does corporate social responsibility affect the institutional ownership of firms in the hospitality and tourism industry?[J]. Tourism Economics, 29(4): 853-879. Marshall A, Rao S, Roy P P, et al., 2022. Mandatory corporate social responsibility and foreign institutional investor preferences[J]. Journal of Corporate Finance, 76: 102261. McCahery J A, Sautner Z, Starks L T, 2016. Behind the scenes: the corporate governance preferences of institutional investors[J]. Journal of Finance, 71(6): 2905-2932. Nath P C, Mishra A K, Sharma R, et al., 2024. Recent advances in artificial intelligence towards the sustainable future of agri-food industry[J]. Food Chemistry, 447: 138945. Nguyen Q P, Vo D H, 2022. Artificial intelligence and unemployment:an international evidence[J]. Structural Change and Economic Dynamics, 63: 40-55. Ocasio W, 1997. Towards an attention-based view of the firm[J]. Strategic Management Journal, 18(S1): 187-206. Panon L, 2022. Labor share, foreign demand and superstar exporters[J]. Journal of International Economics, 139: 103678. Persad K, Xu B, Greening P, 2024. Institutional investor behaviour and the energy transition: a complexity framework for accelerating sustainable finance from UK investors[J]. Energy Economics: 107444. Pinski M, Hofmann T, Benlian A, 2024. AI literacy for the top management: an upper echelons perspective on corporate AI orientation and implementation ability[J]. Electronic Markets, 34(1): 24. Polireddi N S A, 2024. An effective role of artificial intelligence and machine learning in banking sector[J]. Measurement: Sensors, 33: 101135. Qu G, Jing H, 2025. Is new technology always good? Artificial intelligence and corporate tax avoidance: evidence from China[J]. International Review of Economics & Finance, 98: 103949. Raimondo C, 2019. The media and the financial markets: a review[J]. Asia-pacific Journal of Financial Studies, 48(2): 155-184. Rammer C, Fernández G P, Czarnitzki D, 2022. Artificial intelligence and industrial innovation: evidence from german firm-level data[J]. Research Policy, 51(7): 104555. Rees L, Sharp N, Twedt B, 2015. Who’s heard on the street? Determinants and consequences of financial analyst coverage in the business press[J]. Review of Accounting Studies, 20(1): 173-209. Ren B, Qiu Z, Liu B, 2025. Supply chain decarbonisation effects of artificial intelligence: evidence from China[J]. International Review of Economics & Finance, 101: 104198. Roberts D L, Candi M, 2024. Artificial intelligence and innovation management: charting the evolving landscape[J]. Technovation, 136: 103081. Roberts H, Cowls J, Morley J, et al., 2021. The chinese approach to artificial intelligence: an analysis of policy, ethics, and regulation[J]. AI And Society, 36(1): 59-77. Roller S, 2021. Institutional investors’ sustainability policies and the outcomes for the healthcare sector portfolios[J]. Health Policy, 125(10): 1367-1376. Saba C S, Ngepah N, 2024. The impact of artificial intelligence (AI) on employment and economic growth in BRICS: does the moderating role of governance matter?[J]. Research in Globalization, 8: 100213. Saci F, Jasimuddin S M, 2021. Does the research done by the institutional investors affect the cost of equity capital?[J]. Finance Research Letters, 41: 101834. Skare M, Gavurova B, Blažević Burić S, 2024. Artificial intelligence and wealth inequality: a comprehensive empirical exploration of socioeconomic implications[J]. Technology in Society, 79: 102719. Slimani S, Omri A, Ben Jabeur S, 2025. When and how does artificial intelligence impact environmental performance?[J]. Energy Economics, 148: 108643. Song Y, 2024. Institutional investors’ corporate site visits and firm-level climate change risk disclosure[J]. International Review of Financial Analysis. Song Y, Xian R, 2024. Institutional investors’ corporate site visits and firm-level climate change risk disclosure[J]. International Review of Financial Analysis, 93: 103145. Stiglitz J E, 2000. The contributions of the economics of information to twentieth century economics[J]. The Quarterly Journal of Economics, 115(4): 1441-1478. Sullivan Y, Fosso Wamba S, 2024. Artificial intelligence and adaptive response to market changes: a strategy to enhance firm performance and innovation[J]. Journal of Business Research, 174: 114500. Sun Z, Che S, Wang J, 2024. Deconstruct artificial intelligence’s productivity impact: a new technological insight[J]. Technology in Society, 79: 102752. Tao W, Weng S, Chen X, et al., 2024. Artificial intelligence-driven transformations in low-carbon energy structure: evidence from China[J]. Energy Economics, 136: 107719. Teece D J, 2007. Explicating dynamic capabilities: the nature and microfoundations of (sustainable) enterprise performance[J]. Strategic Management Journal, 28(13): 1319-1350. Thompson L, 2025. Human capital disclosures and institutional ownership[J]. Finance Research Letters, 71: 106384. Tian H, Zhao L, Yunfang L, et al., 2023. Can enterprise green technology innovation performance achieve “corner overtaking” by using artificial intelligence?—evidence from chinese manufacturing enterprises[J]. Technological Forecasting and Social Change, 194: 122732. Tie, Y., Liu, Y., 2025. External demand shocks and changes in labor income shares. Finance & Trade Economics,46(7):56-72. Veganzones D, Séverin E, 2025. Earnings management visualization and prediction using machine learning methods[J]. International Journal of Accounting Information Systems, 56: 100743. Wang C, Nazar R, Ali S, et al., 2025. Eco-friendly algorithms: artificial intelligence and green finance in european intelligent nations[J]. Technology in Society: 102960. Wang J, Wang A, Luo K, et al., 2024. Can artificial intelligence improve enterprise environmental performance: evidence from China[J]. Journal of Environmental Management, 370: 123079. Wang L, Chen Y, 2025. Artificial intelligence and corporate investment efficiency: evidence from China[J]. Emerging Markets Review, 68: 101314. Wang Q, Zhang F, Li R, et al., 2024. Does artificial intelligence promote energy transition and curb carbon emissions? The role of trade openness[J]. Journal of Cleaner Production, 447: 141298. Wang S, Sun Z, Wang H, et al., 2025. Enhancing student acceptance of artificial intelligence-driven hybrid learning in business education: interaction between self-efficacy, playfulness, emotional engagement, and university support[J]. The International Journal of Management Education, 23(2): 101184. Wang S, Zhang Y, Xiao Y, et al., 2025. Artificial intelligence policy frameworks in China, the european union and the United States: an analysis based on structure topic model[J]. Technological Forecasting and Social Change, 212: 123971. Wang X, Wei S, Zhu X, 2024. Economic policy uncertainty and heterogeneous institutional investor horizons[J]. Review of Quantitative Finance and Accounting, 62(1): 39-67. Wang Y, Zhang R, Yao K, et al., 2024. Does artificial intelligence affect the ecological footprint? –evidence from 30 provinces in China[J]. Journal of Environmental Management, 370: 122458. Wang, Z., Yang, G., Xiong, H., 2023. Corporate Innovation Focus and Institutional Investor Ownership: Evidence from Listed Firm Patent Classification Number Data. South China Journal of Economics (7): 42-62. Wen H, Shang J, Nghiem X H, 2025. Does artificial intelligence matter for the population aging-inclusive growth nexus? International evidence[J]. Telecommunications Policy, 49(4): 102932. Weng S, Rong Z, Yu L, 2024. Institutional site visits and corporate innovation: the role of analyst coverage[J]. Pacific-basin Finance Journal, 85: 102354. Wernerfelt B, 1984. A resource-based view of the firm[J]. Strategic Management Journal, 5(2): 171-180. Wu B, Chen H, Shi Y, 2025. Influence of artificial intelligence development on supply chain diversification[J]. Finance Research Letters, 78: 107210. Wu Q, Xie T, Lu J, 2025. Does institutional investors’ environmental attention influence corporate environmental investment?[J]. Economic Analysis and Policy, 85: 1625-1644. Wu Y, Lin Z, Zhang Q, et al., 2024. Artificial intelligence, wage dynamics, and inequality: empirical evidence from chinese listed firms[J]. International Review of Economics & Finance, 96: 103739. Wu Y, Zhang J, Cai X, 2025. Impact of regional artificial intelligence development on corporate environmental information[J]. Finance Research Letters, 80: 107413. Xia Y, Xie L, Chen X, et al., 2025. Artificial intelligence and corporate fraud: evidence from China[J]. Economic Analysis and Policy, 86: 1391-1408. Xiao H, 2023. Institutional investors’ corporate site visits and corporate investment efficiency[J]. International Review of Finance, 23(2): 359-392. Xu X, Wan Q, Ma J, et al., 2025. What drives institutional investors’ site visit decisions: the role of local business environment in China[J]. International Review of Financial Analysis, 103: 104204. Yahia N B, Chalwati A, Hmaied D, et al., 2023. Do foreign institutions avoid investing in poorly CSR-performing firms?[J]. Journal of Banking & Finance, 157: 107029. Yan X (Sterling), Zhang Z, 2009. Institutional investors and equity returns: are short-term institutions better informed?[J]. Review of Financial Studies, 22(2): 893-924. Yang Y, An R, Song J, 2025. Impact of enterprise artificial intelligence on social responsibility: evidence from text analysis[J]. Finance Research Letters, 75: 106868. Yang Y, Yi C, Li H, et al., 2025. An analysis on the role of artificial intelligence in green supply chains[J]. Technological Forecasting and Social Change, 217: 124169. Yao, J., Zhang, K., Guo, L., et al., 2024. How does artificial intelligence improve firm productivity? Based on the perspective of labor skill structure adjustment. Management World 40 (2), 101–116. Yuan R, Wen W, 2018. Managerial foreign experience and corporate innovation[J]. Journal of Corporate Finance, 48: 752-770. Zhang C, Yang J, 2024. Artificial intelligence and corporate ESG performance[J]. International Review of Economics & Finance, 96: 103713. Zhang J, 2025. Artificial intelligence and green innovation: investigating the effects of executive pay and firm age[J]. Research in International Business and Finance, 77: 102953. Zhang J, Ye Y, 2023. Does the research meeting affect the shareholding ratio of institutional investors in listed companies? Empirical evidence from China[J]. China Finance Review International, 13(4): 714-733. Zhang P, Ma B, Chi C, 2025. Institutional investors and ESG performance: Evidence from China[J]. Economic Analysis and Policy, 86: 1159-1181. Zhao C, Li Y, Liu Z, et al., 2024. Artificial intelligence and carbon emissions inequality: Evidence from industrial robot application[J]. Journal of Cleaner Production, 438: 140817. Zhao L, Li N, Wu Y, 2023. Institutional investors’ site visits, information asymmetry, and investment efficiency[J]. International Review of Financial Analysis, 88: 102674. Zhong K, Song L, 2025. Artificial intelligence adoption and corporate green innovation capability[J]. Finance Research Letters, 72: 106480. Zhou X, Li G, Wang Q, et al., 2025. Artificial intelligence, corporate information governance and ESG performance: quasi-experimental evidence from China[J]. International Review of Financial Analysis, 102: 104087. Zou J, Zhong X, Gong C, et al., 2025. Examining institutional investor preferences: the influence of ESG ratings on stock holding in China’s stock market[J]. Research in International Business and Finance, 73: 102609. Footnotes International Data Corporate (IDC), founded in 1964, is a global professional consulting services organization affiliated with International Data Group (IDG). International Business Machines Corporation (IBM), headquartered in New York, and founded in the United States in 1911, is the world's largest information technology and business solutions company. The 2017 China Artificial Intelligence Industry Special Research Report states that the surge in the development of artificial intelligence mainly began with the introduction of deep learning algorithms in 2006. Additional Declarations No competing interests reported. Supplementary Files Appendix.docx Cite Share Download PDF Status: Posted Version 1 posted 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-7134593","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":500225682,"identity":"29cd15b4-0e81-4860-9f13-1d0a53239163","order_by":0,"name":"Feng zhao","email":"","orcid":"","institution":"School of Economics, Beijing Technology and Business University","correspondingAuthor":false,"prefix":"","firstName":"Feng","middleName":"","lastName":"zhao","suffix":""},{"id":500225683,"identity":"6346c902-93e1-4c6c-812f-c848428b8789","order_by":1,"name":"Dandan Nan","email":"","orcid":"","institution":"School of Economics, Beijing Technology and Business University","correspondingAuthor":false,"prefix":"","firstName":"Dandan","middleName":"","lastName":"Nan","suffix":""},{"id":500225684,"identity":"b8a681db-a143-4ee8-852c-c7e74f22d76b","order_by":2,"name":"Yao Wu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/UlEQVRIie3NMUvEMBTA8RcCqUPg1oaq9xVeKRRF5L5Kw4FdTrjJqUOn3FK59b6Hi5tIILfcB+jgcCIUxzgIig7mRG5rqJtg/jzy3vIjAKHQnw2rYwBS72Zoc5PxXxJLZbPbgwi2s9RaZOXdSKvH11s9huj66RmqBx/JxAoPL5tYLrKjjU5rvs5PwHS9JG9nmHBkjhCVCKVJHV8wJLX2keyDIy356P6bTIaQ3P1CCw5SiRelpSN06yOTTXd1ukKTNq1UCVHlVHHDoDD9RCymN639rMbRct2Jd3V2vowUtbbqJwAHuD8pdw9zExceABBt9yd5+6HWK0KhUOi/9QXYllRvr34r0wAAAABJRU5ErkJggg==","orcid":"","institution":"School of Economics, Beijing Technology and Business University","correspondingAuthor":true,"prefix":"","firstName":"Yao","middleName":"","lastName":"Wu","suffix":""}],"badges":[],"createdAt":"2025-07-16 01:08:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7134593/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7134593/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":89556418,"identity":"78cc21fd-96e1-4a0e-90b2-24a160a213d9","added_by":"auto","created_at":"2025-08-21 09:31:20","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":155532,"visible":true,"origin":"","legend":"\u003cp\u003eChanges in the annual mean value of artificial intelligence information disclosure\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7134593/v1/c4279638f1270df2ddcd68f9.jpeg"},{"id":92919898,"identity":"32f5ed01-ada1-4538-a64c-ef62529e85bb","added_by":"auto","created_at":"2025-10-07 06:38:58","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2205311,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7134593/v1/909fdd38-8292-435e-9e4d-12e999145e2c.pdf"},{"id":89556417,"identity":"f8238287-ba0d-47cd-8b0f-7367d7d6e90f","added_by":"auto","created_at":"2025-08-21 09:31:20","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":33156,"visible":true,"origin":"","legend":"","description":"","filename":"Appendix.docx","url":"https://assets-eu.researchsquare.com/files/rs-7134593/v1/fd66a28d06c8f3a42af55c06.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"AI Disclosure as an Institutional Investment Catalyst: Evidence from China","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eSince the 18th Party Congress, China has placed significant emphasis on the advancement of artificial intelligence. In 2015, the initiative \u0026ldquo;Made in China 2025\u0026rdquo; explicitly set forth the objective of capitalizing on the opportunity presented by the new wave of technological revolution, accelerating the implementation of development strategies, and establishing a powerful manufacturing nation. Consequently, artificial intelligence has advanced into a phase of swift development, with the artificial intelligence market experiencing continuous expansion. The 2017 \u0026ldquo;New Generation Artificial Intelligence Development Plan\u0026rdquo; constructs the strategic goal of \u0026ldquo;2020 synchronization-2025 breakthrough-2030 leadership\u0026rdquo;. Starting from 2019, a total of 18 pilot zones for the innovation and development of artificial intelligence have been set up in three batches. The 2025 government work report explicitly proposes to continue advancing the \u0026ldquo;artificial intelligence +\u0026rdquo; initiative, and to promote the integration of artificial intelligence with the real economy.\u003c/p\u003e\u003cp\u003eDriven by national policies, artificial intelligence has been widely used in industries such as medicine (Iacucci et al., 2025), finance (Du et al., 2025), education (Wang et al., 2025), and food (Nath et al., 2024), etc. The latest IDC\u003csup\u003e1\u003c/sup\u003e report in 2025 indicates that China will continue to lead the Asia-Pacific region\u0026rsquo;s artificial intelligence development, with total artificial intelligence investment expected to exceed \u003cspan\u003e$\u003c/span\u003e100\u0026nbsp;billion by 2028, at a five-year CAGR of 35.2 percent.\u003c/p\u003e\u003cp\u003eArtificial intelligence technology is profoundly reshaping the architecture of the economy and society, encompassing various key branches such as machine learning (ML), deep learning (DL), and natural language processing (NLP)(Ahmed et al., 2022). Machine learning, deep learning, and other algorithms can help companies to predict financial risks (Veganzones et al., 2025; Lin et al., 2025), and natural language processing can extract information from annual reports that is overlooked by investors, and this potential information can provide investors with investment advice (Cohen et al., 2020). Prior study focuses on the impact of artificial intelligence applications on corporate innovation (Bahoo et al., 2023; Zhong and Song, 2025), macroeconomics (Alonso et al., 2022; Das et al., 2025), and stock forecasting (Chopra et al., 2024). However, there is a question of whether artificial intelligence will affect institutional investor ownership. Against the backdrop of the country\u0026rsquo;s great attention to artificial intelligence, companies have also begun to use artificial intelligence technology to manage their operations, and an increasing number have chosen to include the application of artificial intelligence in their annual reports, such as investments in artificial intelligence and artificial intelligence patents. As a professional investment entity in the capital market, the most significant difference between institutional investors and individual investors is that they focus more on the long-term operation and development of enterprises, and the increase in their shareholding ratio has a positive impact in multiple dimensions. In a way, the disclosure of artificial intelligence information reflects the enterprise\u0026rsquo;s planning and commitment to long-term development. This information often attracts the close attention of institutional investors and then has an important impact on their shareholding decisions. Therefore, it is of great significance to explore the intrinsic relationship between corporate artificial intelligence information disclosure and institutional investor ownership.\u003c/p\u003e\u003cp\u003eBased on data from Chinese A-share listed companies on the Shanghai and Shenzhen stock exchanges from 2007 to 2023, this paper investigates the impact and mechanism of artificial intelligence information disclosure on changes in institutional investor shareholdings. The reasons for selecting China as the research subject are threefold. Firstly, over the past decade, China\u0026rsquo;s artificial intelligence sector has experienced rapid development. The \u0026ldquo;Research Report on China\u0026rsquo;s Artificial Intelligence Regional Competitiveness\u0026rdquo; indicates that the scale of China\u0026rsquo;s artificial intelligence industry has expanded from 10\u0026nbsp;billion in 2015 to 700\u0026nbsp;billion in 2024, sustaining a growth rate of over 20% annually for consecutive years. What\u0026rsquo;s more, IBM\u0026rsquo;s\u003csup\u003e2\u003c/sup\u003e \u0026ldquo;Global Artificial Intelligence Adoption Index 2023\u0026rdquo; notes that approximately 50% of Chinese companies are actively implementing artificial intelligence solutions, encompassing areas such as smart healthcare and smart transportation, thus offering a substantial sample of artificial intelligence applications.\u003c/p\u003e\u003cp\u003eSecondly, in 2004, the State Council issued \u0026ldquo;Several Opinions on Promoting the Reform, Opening-up, and Stable Development of the Capital Market\u0026rdquo;. By 2006, the capital market had experienced rapid expansion, propelling China\u0026rsquo;s institutional investors into a phase of rapid development and establishing a pattern where fund companies, finance companies, insurance companies, banks, QFIIs, and other diverse entities coexist (Jia and Che, 2025). Moreover, domestic financial databases such as Wind and CSMAR offer detailed data on institutional investor shareholdings, with a very high level of data availability, providing a rich data source for this paper.\u003c/p\u003e\u003cp\u003eThirdly, China has become the world\u0026rsquo;s second-largest economy, following the United States, and is also the largest emerging economy. Thus, examining the Chinese model can offer insights not only for governments seeking to enhance capital market laws and regulations but also provide empirical evidence to other emerging economies on managing investment shifts due to artificial intelligence (Song, 2024).\u003c/p\u003e\u003cp\u003eThis paper examines the Management Discussion and Analysis (MD\u0026amp;A) section of corporate annual reports, tallying the frequency of AI-related terminology to create an artificial intelligence information disclosure index. Baseline findings suggest that enhanced artificial intelligence disclosure encourages institutional investors to boost their stakes. This conclusion remains valid after undergoing various robustness tests, including the substitution of dependent and independent variables, increasing clustering levels, excluding municipal samples, and employing alternative estimation methods. To tackle potential endogeneity concerns, the study confirms the reliability of its findings using four distinct approaches: the instrumental variable (IV) method with two-stage least squares (2SLS), the Heckman two-step test, propensity score matching (PSM), and the inclusion of omitted variables. Subsequently, the paper investigates the economic mechanisms by which artificial intelligence disclosure amplifies institutional investor ownership. Corporate artificial intelligence disclosure aids institutional investors in conducting site visits, enhances information transparency, and fosters innovation. Heterogeneity analysis indicates that the positive impact of artificial intelligence disclosure on institutional investor shareholdings is more pronounced in post-2011 samples, among manufacturing firms, in highly competitive environments, and for companies with directors possessing a strong technological background. Further analysis demonstrates that the influence of changes in artificial intelligence disclosure on institutional investor shareholdings is more significant for long-term and stress-resistant institutional investors compared to their short-term and stress-sensitive counterparts, highlighting the varying decision-making processes among different investment entities.\u003c/p\u003e\u003cp\u003eThere are three main contributions in this study. Firstly, prior studies examine firm characteristics (Gaver and Gaver, 1993; Hong et al., 2008), corporate governance level (Chung and Zhang, 2011), market performance (Deb, 2018), macro uncertainty (Huang et al., 2022; Laksmana et al. 2023) and corporate social responsibility (Marshall et al., 2022; Yahia et al., 2023) on institutional investor ownership. Some scholars have also examined the impact of disclosure on institutional investor shareholdings from the perspective of corporate disclosure quality (Cheng et al., 2020). We concentrate on artificial intelligence, thereby broadening the body of research on institutional investor shareholding.\u003c/p\u003e\u003cp\u003eSecondly, we enrich the research on the economic consequences of artificial intelligence. Previous research mostly investigates the impact of artificial intelligence on supply chain (Wu et al., 2025; Yang et al., 2025), productivity (Sullivan and Wamba, 2024; Sun et al., 2024), investment efficiency (Wang and Chen, 2025 ), employment (Saba and Ngepah, 2024), population aging (Wen et al., 2025), environment (Wang et al., 2024), energy (Gao and Wang, 2025 ), ESG (Zhou et al., 2025), green finance (Wang et al., 2025) and green transition (Li et al., 2025). However, there is limited research on the impact of artificial intelligence on institutional investors within the capital market. Consequently, our study aims to address this gap concerning the role of artificial intelligence in the capital market.\u003c/p\u003e\u003cp\u003eThirdly, we reveal that the disclosure of artificial intelligence information contributes to changes in institutional investor shareholdings through three distinct economic mechanisms: institutional investor site visits, information transparency, and corporate innovation. Our study elaborates on the mechanism of action by which artificial intelligence influences institutional investor shareholdings.\u003c/p\u003e\u003cp\u003eThe remainder of the paper is organized as follows. Section 2 reviews the related literature. Section 3 is the theoretical analysis and hypothesis formulation. Section 4 describes our research design. Section 5 reports our empirical results. Section \u003cspan refid=\"Sec45\" class=\"InternalRef\"\u003e6\u003c/span\u003e presents the additional analyses, and Section \u003cspan refid=\"Sec48\" class=\"InternalRef\"\u003e7\u003c/span\u003e concludes.\u003c/p\u003e"},{"header":"2. Literature review","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Artificial intelligence\u003c/h2\u003e\u003cdiv id=\"Sec4\" class=\"Section3\"\u003e\u003ch2\u003e2.1.1 Factors affecting artificial intelligence\u003c/h2\u003e\u003cp\u003eArtificial intelligence is influenced by a variety of factors, and scholars have analyzed it from micro, industry, and macro perspectives.\u003c/p\u003e\u003cp\u003eAt the micro level, data resources, human capital, and technological capabilities all influence the development of artificial intelligence(Pinski et al., 2024). Rich data resources provide ample material for training artificial intelligence models. Professionals can integrate AI technology with the company\u0026rsquo;s core business, and advanced technology aids in the design and optimization of artificial intelligence models. Artificial intelligence represents a long-term investment, and the executive team plays a crucial role in the development strategy and the extent of artificial intelligence application (Li and Shao, 2023). On the one hand, the executive team\u0026rsquo;s artificial intelligence literacy underscores the significance of AI within the organization and motivates the application of artificial intelligence technology to actual business operations (Pinski et al., 2024). On the other hand, executives with a future-oriented perspective can incorporate artificial intelligence into the company\u0026rsquo;s long-term development strategy (Li et al., 2021). In addition, Liu et al. (2025) also found that larger firms with higher firm value are more inclined to apply artificial intelligence.\u003c/p\u003e\u003cp\u003eAt the industry level, within the financial sector, artificial intelligence undertakes tasks such as risk assessment, stock trading, and loan provision, among others. It has changed the way the financial industry provides services to its customers, which can not only improve the efficiency of internal processes, maintain cybersecurity, and reduce financial risks, but can also lead to more personalized products and services (Polireddi, 2024). In the manufacturing industry, the use of artificial intelligence enables real-time monitoring of the production floor, allowing for the identification and prompt resolution of problems. Robots, as automation-oriented applications of artificial intelligence, can replace humans in challenging and hazardous tasks. This not only enhances productivity but also ensures the safety of employees (Lee et al., 2022). Furthermore, industries with a high level of technological intensity are more inclined to invest significantly in artificial intelligence. This includes sectors such as information transmission, software and information technology services, and scientific research and technology services, which possess robust research and development (R\u0026amp;D) capabilities in digital technology. These industries tend to favor the use of cloud computing, big data, and artificial intelligence for managing their companies (Yao et al., 2024).\u003c/p\u003e\u003cp\u003eMacro policy factors: Since 2013, China has released several documents on artificial intelligence, including \u0026ldquo;The New Generation Artificial Intelligence Development Plan,\u0026rdquo; \u0026ldquo;The Guidelines for the Construction of the National New Generation Artificial Intelligence Standard System,\u0026rdquo; and \u0026ldquo;The Guidelines for the Construction of the National Comprehensive Standardization System for the Artificial Intelligence Industry (Version 2024)\u0026rdquo; etc. Roberts et al. (2021) analyzed the nature of China\u0026rsquo;s artificial intelligence policy and the context in which it has emerged, suggesting that there are opportunities for China to develop in artificial intelligence areas such as automation, which will strongly contribute to the growth of the Chinese economy. Wang et al. (2025) analyzed China\u0026rsquo;s artificial intelligence policy using Structural Thematic Modeling (STM) and indicated that China has paid increasing attention to the development of artificial intelligence, focusing on applications in four areas: industrial applications, technical standards, talent training and scientific research. The proposed artificial intelligence policy will encourage enterprises to increase their capital investment in artificial intelligence and patent applications, but weak intellectual property rights will inhibit their investment in artificial intelligence innovation and reduce the incentives for patent development. Therefore, it is also crucial to improve the intellectual property protection system. China has gradually improved its IPR protection system since 2008, and has issued relevant policy documents, such as \u0026ldquo;the Outline of National Intellectual Property Strategy\u0026rdquo;, \u0026ldquo;Measures for the Evaluation of National Intellectual Property Pilot and Demonstration Cities (Districts)\u0026rdquo;, \u0026ldquo;and Outline for the Construction of a Strong Intellectual Property Rights Country (2021\u0026ndash;2035)\u0026rdquo;, etc. Han et al. (2025) employed a time-varying difference model to study the relationship between the intellectual property model city policy and the development of enterprises\u0026rsquo; artificial intelligence which suggests that the policy significantly improves the development level of artificial intelligence.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section3\"\u003e\u003ch2\u003e2.1.2 Economic impact of artificial intelligence\u003c/h2\u003e\u003cp\u003eThe existing literature primarily addresses the economic implications of artificial intelligence at the enterprise operational level, the macroeconomic level, and the energy and environmental levels.\u003c/p\u003e\u003cp\u003eIn terms of enterprise operations, artificial intelligence has been rapidly applied to various fields of the economy and society, reshaping the production and operational modes of enterprises (Culot et al., 2024). Supply chain management: Artificial intelligence technology can provide supply chain services for manufacturing enterprises, aiding in the enhancement of decision-making efficiency and cost reduction through the optimization of inventory management, logistics, and distribution(Berente et al., 2021), and improving supply chain performance (Belhadi et al., 2022). Productivity: Arithmetic power, serving as the cornerstone of intelligent transformation in the era of artificial intelligence, empowers enterprises to extract value and reconfigure the allocation of production factors, thereby enhancing enterprise productivity (Ding and Gao, 2025). The supply chain digitalization transformation policy further enhances the role of AI technology in promoting the productivity of enterprises (Sun et al., 2024). Industrial robotics, as a significant application of artificial intelligence, can enhance the total factor productivity of enterprises and drive their high-quality development (Duan et al., 2023). Czarnitzki et al. (2023) constructed a firm-level Cobb-Douglas production function to further verify the positive relationship between artificial intelligence technology and firm productivity. Labor Income Level: Existing literature indicates that the impact of artificial intelligence on labor income varies by industry and is influenced by policy. In the service sector, artificial intelligence reduces the number of less-educated workers and front-line employees, thereby decreasing the firms\u0026rsquo; share of labor income (Chen et al., 2024). And in the financial sector, Chang (2025) found that artificial intelligence can effectively increase income levels, especially on the East Coast and market-driven regions. Fan et al. (2025) employed industry data to find that next-generation artificial intelligence pilot zone policies increase firms\u0026rsquo; share of labor income through skill demand effects and skill premium effects, which highlights the impact of artificial intelligence policies on labor income.\u003c/p\u003e\u003cp\u003eAt the macroeconomic level, Chishti et al. (2024) considered artificial intelligence as a new driver of the economic cycle. It is documented that AI may cause economic volatility in the short term, yet it has the potential to stabilize the economy in the long term. In terms of economic growth, Saba and Ngepah (2024) examined the impact of artificial intelligence on the economies of the BRICS countries using the Cobb-Douglas production function and the Solow-Swan model. Their results indicated that artificial intelligence has a significant positive impact on the economic growth of the BRICS countries, which is evident in both the short and long term, though it is more pronounced in the short term. Alonso et al. (2022) stated that artificial intelligence may exacerbate the economic divide between developing and developed economies. The economic gap between these economies exists because developed nations utilize more robots, have higher starting wages, greater robot productivity, and experience significantly higher GDP growth rates compared to developing economies. In addition, based on the extended Solow-Swan growth model, Skare et al. (2024) treated artificial intelligence as a form of capital that can replace or supplement traditional forms of labor, and found a significant positive correlation between artificial intelligence capital stock and wealth inequality, which can be explained by the technology substitution effect, that is, the growth of income of high-skilled labor is relatively fast and low-skilled labor income growth relatively slow, leading to wealth-income inequality. However, the increase in capital investment can improve productivity and thus lead to overall economic growth, but it does not change the phenomenon of wealth-income inequality. In the study of artificial intelligence and unemployment, Nguyen and Vo (2022) revealed that artificial intelligence can exacerbate unemployment; however, when inflation reaches a certain threshold, unemployment rates tend to decrease. This suggests that artificial intelligence may help to mitigate unemployment when inflation is at its anticipated level.\u003c/p\u003e\u003cp\u003eAs for Energy and Environment: In recent years, artificial intelligence has been extensively applied in the fields of energy and the environment, facilitating the transition of the economy towards a low-carbon energy structure (Huang et al., 2025). Artificial intelligence can significantly reduce energy consumption by encouraging companies to innovate and drive digital transformation (Fu et al., 2024), improve energy and resource efficiency (Li et al., 2023; Li et al., 2025), and reduce ecological footprints (Wang et al., 2024). In the area of corporate environmental performance, the application of artificial intelligence significantly improves the quality of corporate environmental disclosure (Wu et al., 2025), promotes the growth of environmental performance (Wang et al., 2024), and can effectively transform the growth of environmental performance into the enhancement of corporate reputation and market value (Liu et al., 2025). In the field of carbon emissions, artificial intelligence will reduce carbon emissions by improving production efficiency, optimizing resource allocation, and promoting technological progress (Luqman et al., 2024). Based on cross-country data, Wang et al. (2024) further pointed out that there exists a threshold effect of trade openness on the suppression of carbon emissions by artificial intelligence. When trade openness is below this threshold, the impact of artificial intelligence on carbon emissions is negligible. Conversely, when trade openness surpasses the threshold, artificial intelligence significantly reduces carbon emissions. Furthermore, artificial intelligence can mitigate carbon inequality between nations, and its inhibitory effect on carbon inequality becomes more pronounced as the severity of the inequality increases (Zhao et al., 2024).\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Institutional investor shareholdings\u003c/h2\u003e\u003cdiv id=\"Sec7\" class=\"Section3\"\u003e\u003ch2\u003e2.2.1 Factors affecting institutional investors shareholdings\u003c/h2\u003e\u003cp\u003eIt is well documented that institutional investors consider three levels of factors when making investment decisions: micro, industry, and macro.\u003c/p\u003e\u003cp\u003eAt the micro level, Cheng et al.(2020) indicated that during market downturns, the quality of corporate disclosure exerts a greater positive influence on institutional investor ownership. Institutional investors prefer to invest in listed companies that exhibit faster growth and better profitability (Lin et al., 2014). Because such companies have more resources to develop core products, which provides conditions for them to seize market share. Institutional investors assess a company\u0026rsquo;s growth to ascertain whether its stock price is overvalued or undervalued. If they find the stock to be undervalued, they may purchase shares with the expectation of earning a return as the company expands in the future. Gaver and Gaver (1993) constructed an indicator to assess a company\u0026rsquo;s growth potential, based on six variables, and found that companies on the ascent typically exhibit lower debt-to-equity ratios and dividend yields. These characteristics could potentially draw the attention of institutional investors. Companies with robust profitability also capture investors\u0026rsquo; interest, as these firms often boast stable cash flows and a more secure financial standing. Hong et al. (2008) measured the profitability of a company in terms of ROE, and they verified that it has a significant positive impact on stock prices, thus attracting investors. Corporate Social Responsibility (CSR) Dimension: As the green economy develops, CSR has increasingly become an indicator for evaluating the investment value of companies. Those with strong social responsibility are often more appealing to institutional investors for investment (Lyssimachou and Bilinski, 2023). Li and Lu (2016) even regarded corporate environmental capital expenditure as part of CSR and indicated that environmental capital expenditure promotes the increase of institutional investors\u0026rsquo; shareholding. In addition, the governance level of listed companies determines the efficiency of property allocation and the operation level of enterprises, and institutional investors prefer to hold shares of companies with a good governance structure, and their shareholding ratio will increase with the increase of the quality of corporate governance (Chung and Zhang, 2011).\u003c/p\u003e\u003cp\u003eIndustry Level: The \u0026ldquo;dual-carbon\u0026rdquo; goal is further propelling the development of China\u0026rsquo;s new energy industry, aligning with the global green economy\u0026rsquo;s development trend. The new energy industry encompasses a broad spectrum of sectors, including solar and wind energy, among others. Institutional investors are optimistic about the energy transition, which is expected to encourage them to invest in stocks and benefit from the dividends of this transition (Persad et al., 2024). In the healthcare sector, the growing aging population is driving the expansion of the healthcare industry. Simultaneously, ongoing innovation and breakthroughs in medical technology are providing significant momentum to the industry\u0026rsquo;s development, offering long-term and stable investment opportunities for institutional investors (Roller,2021). Additionally, the technology industry boasts a high level of innovation and growth potential, which can inject continuous vitality into economic development, thereby attracting institutional investors to invest.\u003c/p\u003e\u003cp\u003eAt the macro level, during periods of high political uncertainty, institutional investors utilize the available information to assess the impact of a political event on the market. Consequently, they decide whether to increase or decrease their stock holdings. Francis et al. (2021) stated that institutional investors tend to significantly reduce their common stock positions when political uncertainty is high. Nevertheless, if firms voluntarily disclose information pertaining to political expenditures, this can mitigate the negative impact of political uncertainty on institutional investor ownership (Goh et al., 2020). From the perspective of economic policy uncertainty (EPU), Hao and Li (2024) argued that this uncertainty can render future cash flows unstable, causing institutional investors to adopt a more cautious approach to investing in domestic stocks, thereby reducing their holdings. Wang et al. (2024) further discussed the impact of EPU on investor categories and found that long-term institutional investors react more positively to EPU because they believe that it is a short-term economic volatility behavior that will not affect the long-term value of firms; whereas, short-term institutional investors reduce their stock holdings to cope with market risks. From a global perspective, global volatility also affects the behavior of institutional investors. During periods of high volatility, institutional investors tend to decrease their stock allocations (Kacperczyk et al., 2022). In addition, geopolitical risk (GPR) is also a major factor affecting institutional investor ownership. Choudhury (2025) investigated the impact of GPR on U.S. investors through an extended five-factor model and analyzed the relationship between the GPR factor and the portfolio return, which provides empirical evidence for the adjustment of investment strategies by institutional investors. Furthermore, Fiorillo et al. (2024) demonstrated that GPR enhances institutional investors\u0026rsquo; investment preference for green bonds.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section3\"\u003e\u003ch2\u003e2.2.2 Economic impact of institutional investors shareholdings\u003c/h2\u003e\u003cp\u003eMeanwhile, institutional investors exert a constructive influence on corporate governance, climate risk, and ESG.\u003c/p\u003e\u003cp\u003eIn the realm of corporate governance, institutional investors, as professional participants in the capital market, directly shape the governance framework through their ownership stakes. Appel et al. (2016) challenged the view that passive investors lack governance incentives, revealing that by concentrating their voting power, they significantly improve the corporate governance structure and bolster the long-term performance of firms. McCahery et al. (2016) employed a questionnaire to find that institutional investors are able to influence corporate governance through behind-the-scenes intervention and exit mechanisms. Aggarwal et al. (2011) suggested a positive correlation between institutional investor shareholdings and company valuations, indicating that institutional investor shareholding not only affects corporate governance but also affects company value and board decisions. Moreover, when portfolio company value appreciates, institutional investors can also receive financial incentives such as management fees, which heighten their motivation to participate in corporate governance (Lewellen and Lewellen, 2022).\u003c/p\u003e\u003cp\u003eConcerning climate risk, climate risk poses potential financial risks for institutional investors (Krueger et al., 2019). Site visits, as a crucial engagement channel between companies and investors, inhibit managerial incentives to conceal climate risk information, enhance the quality of the information environment, and compel management to disclose more information about climate change (Song and Xian, 2024). Flammer et al. (2021) argued that institutional investors, especially long-term institutional investors, play an important role in promoting voluntary disclosure of climate risk information. Li et al. (2023) employed the event study method to uncover that climate risk information disclosure can send positive signals to institutional investors, showing significant positive abnormal returns in the samples with high investor attention, which indicates that the market\u0026rsquo;s attitude towards climate risk information disclosure is generally positive. In addition, institutional investors mitigate corporate climate risk exposure by promoting green innovation capabilities (Jin et al., 2024).\u003c/p\u003e\u003cp\u003eInstitutional investors significantly accelerate the development of corporate ESG. Huang et al. (2025) indicated that institutional investors ask ESG-related questions during site visits, which motivates companies to enhance the disclosure of non-financial information on ESG. Analysis of cross-country data by Dyck et al. (2019) confirmed that institutional investor ownership is positively associated with corporate environmental and social (E\u0026amp;S) performance, demonstrating that E\u0026amp;S decisions are not only influenced by managers but also driven by investors. Crucially, Zhang et al. (2025) reported that institutional investor ownership deters corporate greenwash while improving ESG performance, which in turn reduces operational risk and increases firm value. After further categorizing institutional investors, Liu et al. (2023) and Giordino et al. (2025) suggested a positive association between long-term institutional investors and ESG performance.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"3. Theoretical analysis and hypothesis development","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Artificial intelligence information disclosure and institutional investor ownership\u003c/h2\u003e\u003cp\u003eFrom the perspective of information economics, Stiglitz (2000) indicated that imperfect information and information friction costs trigger information asymmetry, which distorts resource allocation and contributes to market inefficiency. The advent of the Fourth Industrial Revolution presents artificial intelligence as a transformative solution to this dilemma. AI technology can effectively break down \u0026ldquo;data silos\u0026rdquo; and reduce the information friction costs of enterprises by enhancing their information acquisition and processing capabilities (Goldfarb and Tucker, 2019; He et al., 2025). Boot et al. (2021) indicated that the AI-powered analysis of unstructured data reveals a significant scale advantage in the financial services sector, effectively bridging the information gaps between financing counterparts. At the corporate level, Begenau et al. (2018) argued that compared to small enterprises, large enterprises attain greater efficiency in AI technology. They can also enter the capital market with lower thresholds, thereby attracting institutional investors. Moreover, companies that disclose more information can enable external investors to assess the firm\u0026rsquo;s value, reducing information asymmetry (Chung et al., 2015).\u003c/p\u003e\u003cp\u003eFrom the Resource-Based View (RBV), Wernerfelt (1984) and Barney (1991) pointed out that resources with value, scarcity, inimitability, and non-substitutability (VRIN)\u0026zwnj; are the prerequisites for a company to have a competitive advantage (Liu et al., 2025a). AI is precisely a VRIN resource (Bag et al., 2021; Li et al., 2025), which belongs to the category of intangible assets (Xia et al., 2025). Among them, value is reflected in the role of artificial intelligence in improving the operational efficiency of enterprises (Slimani et al., 2025); scarcity comes from the three elements of algorithms, arithmetic power, and data; inimitability refers to the complexity of artificial intelligence technology, which is difficult to be copied by competitors; and non-substitutability is reflected in the processing of data and optimization of algorithms. Management can formulate business strategies and participate in corporate governance based on artificial intelligence (Qu and Jing, 2025 ), thus releasing positive signals of long-term stable development to the market, which makes it easier for firms applying artificial intelligence technology to gain the favor of institutional investors and increase their shareholdings.\u003c/p\u003e\u003cp\u003eTherefore, we hypothesize as follows:\u003c/p\u003e\u003cp\u003e\u003cb\u003eH1.\u003c/b\u003e Corporate artificial intelligence information disclosure will prompt institutional investors to increase their stock holdings.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Artificial intelligence information disclosure, institutional investor site visits and institutional investor ownership\u003c/h2\u003e\u003cp\u003eThe opacity of artificial intelligence\u0026rsquo;s algorithmic black boxes has exacerbated social trust complexities, particularly undermining institutional investors\u0026rsquo; confidence in AI-related information disclosure in their annual reports. To explore the use of artificial intelligence and verify the authenticity of artificial intelligence technology, institutional investors increasingly employ site visits as a due diligence mechanism, which can not only observe the company\u0026rsquo;s R\u0026amp;D circumstance, grasp the production status and development plans (Wu et al., 2025), but can also obtain the information not disclosed in the annual reports (Zhao et al., 2023), understand the company\u0026rsquo;s prospects and business risk exposure (Jiang and Yuan, 2018), and then make investment decisions to influence the operation of the capital market (Belev, 2003; Saci \u0026amp; Jasimuddin, 2021; Xiao, 2023; Zhao et al., 2023). Furthermore, institutional investors frequently convene meetings following site visits, during which they engage in detailed discussions with management. These meetings also furnish institutional investors with more reliable information, influencing their decisions on stock ownership (Zhang \u0026amp; Ye, 2023). Furthermore, following the site visits, company management will proactively enhance the quality of information disclosure to prevent an increase in the degree of information asymmetry between individual and institutional investors, thereby improving stock liquidity (Balakrishnan et al., 2014) and promoting the increase of institutional holdings.\u003c/p\u003e\u003cp\u003eThen, we hypothesize as follows:\u003c/p\u003e\u003cp\u003e\u003cb\u003eH2.\u003c/b\u003e Artificial intelligence information disclosure can promote institutional investor shareholdings by increasing the frequency of institutional investors\u0026rsquo; field visits.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Artificial intelligence information disclosure, information transparency and institutional investor ownership\u003c/h2\u003e\u003cp\u003eAccording to the Attention-Based View (Ocasio, 1997), corporate disclosure of artificial intelligence information can be viewed as a conscious attention-guiding behavior. This selective disclosure strategically directs organizational focus toward AI-related matters, thereby influencing the salience allocated to such information within the firm and ultimately reshaping the attention distribution pattern (Joseph and Wilson, 2018).\u003c/p\u003e\u003cp\u003eAs an information intermediary, the media, with its unique perspective on corporate information, presents a dual effect: it not only strengthens the transparency of corporate information, thereby increasing social attention, but also plays a supervisory role in corporate governance (Raimondo, 2019). By reporting positive news, such as the advantages that AI applications bring to a company\u0026rsquo;s operations, the media can motivate companies to focus more on the research and development of artificial intelligence. Conversely, if the press highlights the disadvantages of AI technology, such as data breaches, this can compel companies to disclose more information about their artificial intelligence initiatives to rebuild their corporate image and salvage their reputation. Whether it is a positive or negative report, the press can shape the information environment of a firm (Bushee et al., 2010) and increase information transparency by packaging and disseminating information. Institutional investors are able to get more comprehensive information through these news reports, which is beneficial to their investment decisions.\u003c/p\u003e\u003cp\u003eAnalysts, as professional information decoders (Rees et al., 2015), are an important part of the capital market, which can reduce the degree of information asymmetry between the company and investors, and encourage investors to invest. Among the various types of investors, institutional investors are closely linked to analysts. In the context of the increasing complexity of the capital market, institutional investors are increasingly relying on analysts\u0026rsquo; interpretation of data (Zou et al., 2025). The advent of artificial intelligence has elevated this dependency to a new level. With the assistance of AI technology, analysts can more accurately evaluate a company\u0026rsquo;s profitability, management, and sustainable growth, and then compile these findings into professional analysis reports. Consequently, institutional investors can utilize these reports to assess a company\u0026rsquo;s fundamental status and subsequently determine whether to increase or decrease their shareholdings (Kong et al., 2021).\u003c/p\u003e\u003cp\u003eThis synergy between the media and analysts enhances the transparency of corporate information and reduces the cost of obtaining corporate information for institutional investors (Armstrong et al., 2011), thus providing strong support for their investment decisions.\u003c/p\u003e\u003cp\u003eThus, we hypothesize as follows:\u003c/p\u003e\u003cp\u003e\u003cb\u003eH3.\u003c/b\u003e Corporate artificial intelligence information disclosure can increase institutional investor ownership by enhancing information transparency.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e3.4 Artificial intelligence information disclosure, innovation and institutional investor ownership\u003c/h2\u003e\u003cp\u003eIn the digital age, innovation is influenced by factors such as technological change, economic development, and human capital (Haefner et al., 2021). The nature of innovation is a learnable process of \u0026ldquo;generating new ideas - creative thinking - commercialization of services\u0026rdquo; (Bahoo et al., 2023), which provides long-term value to investors through the creation of competitive advantage and the ability to serve new markets (Sakaki \u0026amp; Jory, 2019).\u003c/p\u003e\u003cp\u003eCorporations\u0026rsquo; growth prospects are a factor influencing institutional investor shareholdings, and innovation capability, as a major assessment indicator of firms\u0026rsquo; growth, not only reflects firms\u0026rsquo; ability to perceive technological opportunities, but also reflects the potential for resource reconfiguration and long-term growth (Teece, 2007; Gao et al., 2025). Innovation, as a signal independent of short-term financial performance, provides a new perspective for institutional investors to assess the quality of management and technological potential, which directly influences their stock holding decisions (Aghion et al., 2013). In this process, artificial intelligence can be used as an aid to R\u0026amp;D to unleash the innovation potential (Johnson et al., 2022; Roberts and Candi, 2024). Artificial intelligence can enhance innovation by increasing digital adaptability (Gao et al., 2025), improving the efficiency of firms\u0026rsquo; operations (Rammer et al., 2022), and facilitating the efficiency of operations.\u003c/p\u003e\u003cp\u003eMoreover, the enhancement of enterprise innovation ability releases the signal of long-term growth to institutional investors. Wang et al. (2023) indicated that the stronger the innovation ability of the company, the higher the proportion of institutional investor shareholding.\u003c/p\u003e\u003cp\u003eAccordingly, we hypothesize as follows:\u003c/p\u003e\u003cp\u003e\u003cb\u003eH4.\u003c/b\u003e Artificial intelligence information disclosure will increase institutional investor ownership by improving the corporate innovation level.\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Research design","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Model specification\u003c/h2\u003e\u003cp\u003eTo study the impact of corporate artificial intelligence information disclosure on the change in institutional investor ownership, we establish the following model:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{dIO}_{i,t}={\\beta\\:}_{0}+{\\beta\\:}_{1}{dAI}_{i,t}+\\sum\\:{\\beta\\:}_{j}Control{s}_{i,t}+{\\eta\\:}_{i}+{\\sigma\\:}_{t}+{\\epsilon\\:}_{i,t}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eIn Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), \u003cem\u003ei\u003c/em\u003e represents the firm, \u003cem\u003et\u003c/em\u003e represents time, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{dIO}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e is the change value of the dependent variable, institutional investor ownership, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{dAI}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e is the change value of the core explanatory variable, artificial intelligence information disclosure index; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:Control{s}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e is a series of control variables; \u003cem\u003e\u0026eta;\u003csub\u003ei\u003c/sub\u003e\u003c/em\u003e represents firm fixed effects, \u003cem\u003e\u0026sigma;\u003csub\u003et\u003c/sub\u003e\u003c/em\u003e represents year fixed effects, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\epsilon\\:}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e is the random error term. Additionally, we also cluster standard errors at the firm level.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Variable selection\u003c/h2\u003e\u003cdiv id=\"Sec17\" class=\"Section3\"\u003e\u003ch2\u003e4.2.1 Dependent variable\u003c/h2\u003e\u003cp\u003eThe change value of institutional investor ownership (dIO): Following Bai et al.(2022), the institutional investor ownership constructed in this paper is the sum of the total shareholding ratio of institutional investors in listed companies, including the shareholding ratio of funds, QFIIs, brokers, insurance companies, social security funds, trust funds, finance companies and banks.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section3\"\u003e\u003ch2\u003e4.2.2 Independent variable\u003c/h2\u003e\u003cp\u003eThe change value of the artificial intelligence information disclosure index (dAI): We construct the artificial intelligence information disclosure index based on the MD\u0026amp;A section in annual reports with the method of Zhang (2025) and Yao et al. (2024).\u003c/p\u003e\u003cp\u003eFirst, we collected annual reports of listed companies from 2007 to 2023, and employed regular expressions to match keywords such as \u0026ldquo;Board of Directors Report\u0026rdquo; or \u0026ldquo;MD\u0026amp;A\u0026rdquo; at the beginning and \u0026ldquo;Important Matters\u0026rdquo; at the end to accurately extract the contents of the MD\u0026amp;A section in the annual report.\u003c/p\u003e\u003cp\u003eSecondly, we manually selected 52 terms, including \u0026ldquo;Artificial Intelligence,\u0026rdquo; \u0026ldquo;Machine Learning,\u0026rdquo; \u0026ldquo;Internet of Things,\u0026rdquo; and others, as seed words. Next, we utilized the text from the annual report as the training corpus for the Word2vec technique and the Skip-gram model. Based on the cosine similarity between the seed words and the output words, we filtered the 10 words most semantically similar to each seed word. Subsequently, we removed duplicate words, words unrelated to artificial intelligence, and words with very low frequency. We ultimately obtained a list of 73 words as the artificial intelligence lexicon for this paper.\u003c/p\u003e\u003cp\u003eFinally, \u0026ldquo;jieba\u0026rdquo; is used to segment the content of the MD\u0026amp;A. The artificial intelligence lexicon is added to the \u0026ldquo;jieba\u0026rdquo; segmentation module as a predefined proper noun. The frequency of occurrences of artificial intelligence keywords in the segmentation is then counted, and this count is used as the artificial intelligence information disclosure index.\u003c/p\u003e\u003cp\u003eTo further alleviate the endogeneity problem, this paper differentiates the dependent variable and the independent variable to eliminate the impact of fixed effects that do not change over time (Jiang et al., 2011; Panon, 2022; Thompson, 2025; Tie \u0026amp; Liu, 2025). Meanwhile, to facilitate the economic meaning of the coefficients, we standardized both the dependent variable and the independent variable, resulting in dIO and dAI (Babina et al., 2024).\u003c/p\u003e\u003cp\u003eAs can be seen from Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, the development of artificial intelligence in China has shown a significant upward trend since 2007, and the fastest growth occurred in 2020. Maybe this acceleration is driven by the COVID-19 pandemic, which not only accelerates the use of artificial intelligence in healthcare (Annan and Qingge, 2025) but also forces enterprises to restructure their production mode as their employees adopt remote work to maintain business operations. Artificial intelligence has provided critical technological support for this transition (Aleem et al., 2023).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section3\"\u003e\u003ch2\u003e4.2.3 Control variables\u003c/h2\u003e\u003cp\u003eTo avoid the problem of omitted variable bias, this paper refers to Li et al. (2016) and Zhang et al. (2023) to select control variables that may affect institutional investor ownership, ensuring an accurate assessment of the disclosure of artificial intelligence. Specifically, the following three types of variables are included: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) corporate characteristics: firm size (Size), financial leverage (Lev), return on total assets (Roa), fixed assets ratio (Mortgage), market-to-book ratio (MB), and firm listing tenure (Age); (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) management characteristics: the percentage of independent directors (Independence), and shareholding ratio of the largest shareholder (Top1); (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) market performance aspects: annual turnover of outstanding shares (Turnover), and annual stock return (Return).\u003c/p\u003e\u003cp\u003eThe definitions of specific variables are shown in the appendix.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Sample source\u003c/h2\u003e\u003cp\u003eThis paper takes 2007 as the starting year\u003csup\u003e3\u003c/sup\u003e and selects the data of China\u0026rsquo;s A-share listed companies in Shanghai and Shenzhen stock exchanges from 2007 to 2023 as the research sample. Then, we process the data as follows: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) removing financial enterprises; (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) excluding samples of ST、PT and *ST enterprises; (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) dropping missing values; (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) winsorizing continuous variables at the 1% and 99% level; (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) deleting information transmission, software and information technology service industry and scientific research and technical service industry.\u003c/p\u003e\u003cp\u003eWe collect financial data from the China Stock Market and Accounting Research (CSMAR) database. The institutional investor ownership data originates from both the Wind database (Wind) and the CSMAR database. Additionally, we obtain stock turnover rates from the RESSET database.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\u003ch2\u003e4.4 Descriptive statistics analysis\u003c/h2\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the descriptive statistics for the primary variables. For Institutional investor ownership (IO), the minimum and maximum values are 0.000 and 0.665, respectively, indicating a significant gap between the extreme values. The mean and median values are 0.064 and 0.036, respectively, with a standard deviation of 0.077. This suggests that institutional investor ownership is generally low across most firms, with only a select few able to secure substantial institutional investment.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eDescriptive statistics\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMean\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSD\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMin\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eMedian\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMedian\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.064\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.077\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.036\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.665\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e4.353\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e11.804\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e430.000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSize\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e22.365\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1.296\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e19.858\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e22.182\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e26.220\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLev\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.451\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.199\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.063\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.449\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.896\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRoa\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.038\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.061\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e-0.199\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.035\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.214\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMortgage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.232\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.163\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.002\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.202\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.719\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMB\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.636\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.253\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.128\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.632\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1.185\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2.362\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.641\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.693\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e2.485\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e3.367\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIndependence\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e37.417\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.395\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e27.270\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e33.330\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e57.140\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTop1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.346\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.148\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.089\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.324\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.749\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTurnover\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e5.872\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.900\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e3.362\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e5.966\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e7.722\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eReturn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.175\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.615\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e-0.668\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.019\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e2.791\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eRegarding AI, the minimum and maximum values are 0 and 430, respectively. The mean and median values are 4.353 and 1, both significantly lower than the maximum value. This indicates that there are substantial disparities in corporate artificial intelligence information disclosure, with only a few companies inclined to disclose more information about artificial intelligence.\u003c/p\u003e\u003cp\u003eAll control variables exhibit data characteristics within optimal ranges, show no signs of outlier contamination, and are consistent with theoretical expectations.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\u003ch2\u003e4.5 Correlation analysis\u003c/h2\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the Pearson correlation coefficients for all variables, none of which exceed the threshold of 0.6, indicating that there is no evidence of severe multicollinearity among the explanatory variables.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eCorrelation analysis\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"13\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c12\" colnum=\"12\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c13\" colnum=\"13\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eIO\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAI\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSize\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eLev\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eRoa\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMortgage\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eMB\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c9\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c10\"\u003e\u003cp\u003eIndependence\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c11\"\u003e\u003cp\u003eTop1\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c12\"\u003e\u003cp\u003eTurnover\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c13\"\u003e\u003cp\u003eReturn\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.002\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSize\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.189***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.029***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLev\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.004\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.063***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.443***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRoa\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.294***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.004\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.047***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.365***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMortgage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.038***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.155***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.045***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.046***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-0.070***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMB\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.193***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.076***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.558***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.370***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-0.221***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.096***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.021***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.097***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.373***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.273***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-0.147***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.028***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.221***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIndependence\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.014**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.045***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.022***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.013**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-0.026***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-0.053***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-0.021***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-0.024***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTop1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.084***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.084***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.210***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.046***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.134***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.076***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.134***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-0.048***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e0.035***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTurnover\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.098***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.015***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.393***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.025***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-0.075***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-0.260***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-0.199***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e-0.039***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e-0.160***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c13\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eReturn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.144***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.033***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.091***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.00200\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.140***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.035***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-0.340***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-0.072***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e-0.029***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e0.013**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c12\"\u003e\u003cp\u003e0.329***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c13\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"13\"\u003eNote: This table reports the Pearson correlation coefficients between the main variables in the study.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"5. Empirical results","content":"\u003cdiv id=\"Sec24\" class=\"Section2\"\u003e\u003ch2\u003e5.1 Baseline regression results\u003c/h2\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents the results of the baseline regression. To ensure the robustness of the regression results, column (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) shows that dAI is positively related to dIO at the 1% significance level, without any control variables included. Columns (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e), and (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) report the regression results with the control variables added sequentially in terms of firm characteristics, management characteristics, and market performance. The coefficients of dAI remain positive and statistically significant. Considering column (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) as the benchmark, the coefficient of dAI is 0.013 and is significantly positive at the 5% level, indicating that a one-standard-deviation increase in dAI corresponds to a 0.013-standard-deviation increase in dIO. This supports Hypothesis 1, which posits that increased disclosure of artificial intelligence will prompt institutional investors to increase their shareholdings.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eBaseline regression results\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003edIO\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.015***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.014**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.014**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.013**\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(2.647)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2.483)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(2.484)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(2.449)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSize\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.010\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.010\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.069***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(-0.778)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(-0.761)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(-5.598)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLev\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.064\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.065\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.068\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(1.061)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(1.072)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(-1.191)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRoa\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.347***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.351***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.459***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2.694)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(2.722)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(-3.631)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMortgage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.162**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.162**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.251***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(-2.457)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(-2.459)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(-4.001)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMB\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.836***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.835***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.135***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(-18.199)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(-18.089)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(2.956)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.342***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.346***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.280***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(-11.611)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(-11.427)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(-9.894)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIndependence\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(-0.798)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(-0.798)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTop1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.044\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.254***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(-0.551)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(-3.305)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTurnover\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-0.139***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(-12.920)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eReturn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.808***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(37.268)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e_cons\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.014***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.549***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.610***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.019***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(-349.254)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(5.591)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(5.686)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(10.787)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj.R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.054\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.037\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.038\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.058\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"5\"\u003eNote: This table reports the results of benchmark regressions of change in artificial intelligence information disclosure (dAI) on change in institutional investor shareholdings (dIO). Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e\u003ch2\u003e5.2 Robustness analysis\u003c/h2\u003e\u003cdiv id=\"Sec26\" class=\"Section3\"\u003e\u003ch2\u003e5.2.1 Replacement of the independent variable\u003c/h2\u003e\u003cp\u003eDrawing on (Yao et al., 2024), we further analyzed the text of annual reports from listed companies, extracted the artificial intelligence word frequency based on the entire annual report data, constructed the indicator AI_words, and incorporated it into Eq.\u0026nbsp;(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) as the replaced explanatory variable to re-estimate. The regression results are shown in column (1) of Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, where the coefficient is 0.013, significant at the 5% level. This finding suggests that our results are robust to this alternative proxy for the artificial intelligence information disclosure index.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eRobustness analysis\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003edIO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003edIO1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003edIO2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003edIO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003edIO\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI_words\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.013**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(2.045)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.017***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.012*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.013***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.013**\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2.815)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(1.679)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(2.723)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(2.242)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e_cons\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.954***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.366\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.077***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.019***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e3.161***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(9.931)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(1.005)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(5.853)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(7.852)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(10.022)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e31352\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31318\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e31318\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e26304\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj.R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.057\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.044\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.035\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.058\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.060\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"6\"\u003eNote: This table reports the regression results of the robustness tests. Column (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) replaces the independent variables. Columns (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) and (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) replace the dependent variables. Column (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) increases the clustering hierarchy to industry, and Column (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) removes municipality samples. In particular, dAI _words, dIO1, and dIO2 are variables obtained by differencing and standardizing the frequency of extracted artificial intelligence words based on annual reports, the ratio of the number of shares held by institutional investors to the total number of shares at the end of the year, and the ratio of the number of shares held by institutional investors to the number of shares outstanding at the end of the year. Code FE denotes firm fixed effects, and Year FE denotes year fixed effects. Standard errors are clustered at the firm level, and values in parentheses are t-values. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec27\" class=\"Section3\"\u003e\u003ch2\u003e5.2.2 Replacement of the dependent variable\u003c/h2\u003e\u003cp\u003eFirstly, we follow the method of Li et al. (2016) and construct the ratio of the number of shares held by institutional investors to the total number of shares (IO1). Secondly, we follow the method of Zou et al. (2025) and construct the ratio of the number of shares held by institutional investors to the number of shares outstanding (IO2). Both variables are reintroduced as dependent variables in Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The regression results are presented in columns (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) and (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, and the coefficients for dAI remain positive and statistically significant (0.017 at 1% level and 0.012 at 10% level, respectively). These findings strengthen the support for Hypothesis 1, which posits a positive impact of artificial intelligence information disclosure.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec28\" class=\"Section3\"\u003e\u003ch2\u003e5.2.3 Changing the clustering level\u003c/h2\u003e\u003cp\u003eWe then shift the clustering level from the firm to the industry to control for factors that cannot be captured at the individual level. The subsequent regression results (column 4, Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) indicate that the coefficient of dAI is 0.013, which is significantly positive at the 1% level, suggesting that the conclusion remains robust after upgrading the clustering level.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec29\" class=\"Section3\"\u003e\u003ch2\u003e5.2.4 Deletion of samples\u003c/h2\u003e\u003cp\u003eConsidering that the special administrative status and city attributes of China\u0026rsquo;s four municipalities directly under the central government (Beijing, Tianjin, Shanghai, and Chongqing) might result in atypical estimates, we follow Tian et al. (2023) to rerun the model after excluding companies from those places in the robustness analysis. As shown in Column (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the coefficient for dAI is 0.013 and significant at the 5% level, confirming the robustness of our primary findings.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec30\" class=\"Section3\"\u003e\u003ch2\u003e5.2.5 Quantile regression\u003c/h2\u003e\u003cp\u003eRecognizing that OLS regression can only capture the average impact while there is a large gap between firms, we adopt the quantile methodology pioneered by Koenker et al. (1978). We estimate at the 10%, 25%, 50%, 75%, and 90% quantiles controlling for firm and year-fixed effects (Agnese et al., 2024). As shown in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, the regression coefficients for dAI are 0.020, 0.013, 0.009, 0.010, and 0.029 for each quartile, respectively. All are significantly positive at the 5% or 1% level, which strongly supports the robustness of the baseline regression and suggests it is not influenced by extreme values. Furthermore, it can be observed that the coefficients exhibit a U-shaped pattern, with an initial decline followed by an increase. This indicates significant heterogeneity in the positive effect across institutional investor shareholdings.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eQuantile regression\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e10%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e25%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e50%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e75%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e90%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.020**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.013**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.009***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.010**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.029***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(2.017)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2.518)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(3.744)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(2.097)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(4.117)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e32267\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32267\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e32267\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e32267\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e32267\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"6\"\u003eNote: This table reports the regression results of the robustness tests, which are quantile regressions with a change in estimation methodology. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, and values in parentheses are z-values. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec31\" class=\"Section2\"\u003e\u003ch2\u003e5.3 Endogeneity tests\u003c/h2\u003e\u003cdiv id=\"Sec32\" class=\"Section3\"\u003e\u003ch2\u003e5.3.1 Instrumental variable approach and 2SLS estimation\u003c/h2\u003e\u003cp\u003eAlthough the baseline regression includes a series of variables that may affect institutional investors\u0026rsquo; shareholding, there are still some unobservable factors related to the level of artificial intelligence that could lead to errors in the results of the benchmark regression. Therefore, we employ a two-stage least squares regression to mitigate the endogeneity issue.\u003c/p\u003e\u003cp\u003eFirstly, following Zhang et al. (2024), we use the lagged period of the independent variable (L.dAI) as the instrumental variable. This instrument is correlated with the current period of the independent variable but is not correlated with the error term of the current period, thus satisfying both correlation and exogeneity.\u003c/p\u003e\u003cp\u003eIn addition, we also construct a Bartik instrumental variable regarding Goldsmith-Pinkham et al. (2020) and Acemoglu et al. (2022), which is the cross-multiplier of the difference between the firm\u0026rsquo;s artificial intelligence and the mean of the industry\u0026rsquo;s annual artificial intelligence with the exclusion of the firm\u0026rsquo;s data\u0026zwnj; and the lagged one-period term of the artificial intelligence, as shown in Eq.\u0026nbsp;(2).\u003c/p\u003e\u003cp\u003e\u003cimg src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAmsAAAA9CAYAAADyBC4nAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAAFiUAABYlAUlSJPAAACDmSURBVHhe7d13VBRn9wfwL8siVRBRQUWlWAALYBAJlhgFIUaxK/b4agQ1sSf60yS2vIkliTV2jRoNFmyJYosV7FQbIL333svu/f1BUQYWQSFZfe/nnD0nzH1mIDOz7t07T1EgIgJjjDHGGJNLIuEGxhhjjDEmPzhZY4wxxhiTY5ysMcYYY4zJMU7WGGOMMcbkGCdrjDHGGGNyjJM1xhhjjDE5xskaY4wxxpgc42SNMcYYY0yOcbLGGGOMMSbHFHgFA/beK4zCRbfzeBSTJoxU0aKrIyaNsoaWgjDCGGOM/Xs4WWPvv4wbWOA0G1u8goWRKkwn7MDFI7PRgevNjDHG5Agna+z9Jy1GTmY28kskwkgVYpWmaKapBkWurDHGGJMjnKwxxhhjjMkxfuDDGGOMMSbHOFljjDHGGJNjnKwxxhhjjMkxTtYYY4wxxuQYJ2uMMcYYY3KMkzXGGGOMMTnGyZpckEIiqX0OMPaOICkkEqlwK2OMMfbGOFn7VxUixv8iNi2ZgEUHbqJYGGbvnqIgHJg3H8t/OYGncXxFGWPsXRQQEIANGzYIN9fJihUrEBAQINz8Vhp8UtyUwBs4ffsFRKKyPJCIYGw7HIO66YJSn+PcVS+k5L6cIl6jhRWGjLSEVvnP+TEB8Lj+ABnFZW2kUilamA7C6P4dK/d5L+SGwf3X/8OuP6PRepAzXKe7oI+hqrBVNdlRPvjrii80Le0xxMoAisIGAJKfX8cZz5DKayCVStG530h8bNZK2PQ9UoKwu1dx60UOLOyc0FNfxrksisXf7lcQkf+yktm0tQUGf9oLzRtk5YIixFx1w77te/FnpAEmb1qPeQP1oSRs9i/KDLuDU38/hfSV96h+z0/wiVU7NMgpEEgNugsPz2fQsxmKwd1bC8MACDE+l3DRJwYKChXvexG62Y1GH+NmwsaMMdaojh07htu3b2PHjh1Vtnt4eGDlypXw9vYGAFhZWWH16tUYMmRIlXZpaWn49ttv0b9/fzg7O1eJvTFqYKnBnrRt9ifUSgEENKG+Q36ga8+TyoNBtHf+cGqlpU4ASNPUiTYc96esV/bPj31EJ7+1IXUFEAAyHDqdfvIKe6XFeyAnlHa6WJGe/hDaeDWMpMK4LJJs+m12F1IBSHvSdiqQsWNK0G3a6upALcqvQb9P19GNoBRhs/dKccwl+tpOjxS1zGjSgafC8EuFcXTtwEpy7KRKAKhF609p3RFfShe2e1tZoXRh9UDqYGNPS84ECqP/qszw+7R7/gDSVwIBimRj9x395RtT9/uwPooSaPeMTiSGIhnMOSKMlpNSrN8VWj/9I9JUAAEaZDdmK90PzxQ2ZIyxRuXm5kYODg7CzbR+/XoCQFZWVuTg4EDa2toElOUpu3btEjan1NRUsrKyIjc3N2HojTR4skZElOW1nqy0Qa16jKYLEYJgXgh949iOAEUatvZyzR8QUW70sY4qaX0whU4FZguj77gi8lhtT+panenrP4KEwVolX9lARpoiEqmAdMb+QmmlwhYvZd7+gXpqg/Qsx9HFSGH0fZNLxxYMJmUxCC2NaPwOb2EDgXha7ahCgBrN2P66tm8jldznW1Hzjg50xLue97GkWLhFJmndm1Yq8NtGH+uBNIz60++PhdGGE3Z8GbVUBomUxWQ4cz8VChu8Iv78YjJWAxn0m0We5d/vGGPsn+Lv708AKCysaoEoLCyMrKysyNPTs3JbamoqzZ49uzJhE+5DrxzP399fGKq3RumzFhuRiJBMoEvv8ehtIAiqtYauThGAFjBs16rGxy7Jj/0RqaCLaXOXY5RJU2H4nZYW4IZt++6g+5QVWDahizAsW9Yz/HHyGtraTEMfY0CUUYS0UmGjl2IjkhCSCZjYjEfvDsLo+yXl3i788UAJTiPtoVkUCYWi2p/sFye/QHxECaDeBwMHNebjdR2MWroOH+Vex/pdxxFbxy5sOWnXsXr+NCzZ64USYVAgPWAPZk12xdGH6cJQrWLCkhGSBLTr8Sl61uM2rA9JihcOnfKBud1UWLQrhWJ2KWT/lRLERaQjOh+wGOAMq/f5iT1jTC7NnDkT48ePh5GRUZXtV69exb59+9C3b9/KbTo6OtixYweMjY0BAO7u7q/sUcbc3BwODg6YOXOmMFRvjZCs5SEmLA5ZJIaofTM0F4ZJhKbaegCykJKWgWr5Rl4wdh66DE17V8yfZCKMvuOK8fDoDlzM6oShnwyCtjAsUyG8dq3BtkADrFgzEW2LgZLydL5meYgOi0MOKUGk3wzvda+f3KfY9N0+ZPSficXjukOUI8XrxmIWJIXjRYQE6NIBBloawnCDUmg9AJMmm+OJ+w6c9IoVhmsmJZQ8u4Wfl8zCcrfHMq9z5uPT+GLKNzjnkwulJsJobUoQFxGFFAJUOnWGSb32rassXNq0Dn9k9cI3y4ejRR5QQrXds1mICElAiUJTiPU0oSIMM8ZYIzp27Bi8vb0xYsQIYQguLi4wNzcXbgYAzJo1S7ipipEjR8Lb2xvHjh0Thuql4ZO14hQ8CQ8ElFuhr2XX6pUzBRHU1JsBKEROXl61ZM3n6DrsvayE6fNnwahRPkT+RZneuOKZgPbm1ujXq40wKlNe8EX8ei4QdjOXws5QGSQGJJEpSCuV8dFXnITHEUGAqi76WJpWvwbvDSl83bfiZn5PfPXFMOgpABpiReRGJiNH2PQVyc/94VsC9PzAAkatahqi0ZAUYWE3EiaZfvjdwwt5wnANmrYchNV7duO7nhn4ecYoLHV7AuHELpkBx7Bg8lScS++FHw9twziLal+LZJOm43nIMxQpaKCLQetG+EcASPN2x74bKRg/bzH66BLEKkBJRDLSZdyylBsLv+hQKOgY4kOLRir1McaYDAcPHgQAWFtbC0N10q1bN+EmAICNjQ0A4OzZs8JQvTT4v9PF8SHwf/QESmpW6GGmIwwDEEFVVa3sv0SiKolEUfRVbN5+BR1mL8Ps3nWvO7214ix4Xz6Mw4fr9/r99yO4/iRBeDSZkh9exemnKehhMxqWLeuaQuXg5H+/wW2dMVjhbAjFInUoKgGSvDRk5go/wssUx76A/6OnUFbrhR6m9fgQf8dIEm5g265rEH00EcP0FVGQXwolkRR52Rm1JEVSPPe7gyxSxAfdP4BOY+dqAHT1O8O4DRBy6wlC8mVkKwKKHYfi2137MMI8GxvnfY4F515Uxij+LBZ8PhvuadZYe2g7ZvSp33tFkhQOPx8fQKknBtl2FoYbQBIOrfkePm3GYNmQ5hAVqkNRDJTkpSEzX9i2TF74U/h7h6K5tjW6mzRutZMxxl6VlpaGy5cvA0C1R6CvExkZCW1tbfTu3VsYAsofhQLA8ePHkZaWJgzXnbAT29uKuLyOujYBqdkupxf5EmGYiCTktXUyAaDh35+ngsrtpeQ235KMek6ny9H5VfZodOlPaYVdWSfB+rxEYmWyXf6X8GgyFJPP5v+QGGo0ausDYVCmmEuryM6mL60+F1W2If4BufbTJHHLT+iPoJq7a4d6/JdMlEAafb+lEFlDRt95+XRhlT0Z9ppGN+PLethH/7WcTJspkdX4zRQibF4h/ymtG9aFoN2V1vz5z4y8kCbepy8+1iSl5iPJPaLmayZLYuhf5NS3DaGZBe24Ek3Fkddp5XBdErexoO/OJtAbjC2g9CdHaLA6SNRxJt1IKhKG39pzty+pt9VA+uXvRCIiKg29RBM/0CBN44l0PlbYmohIQoEnV1AbgNo6bSIeW8AY+yd5enoSykd61kdqaippa2vXOBr0VRUjR18doFBfDVxZkyLqmQ8Ci0Uw+9AMBqo1H14kUoQYQEJKauXjnbx7O7HaPR2DFy3A4HYy5shqLJodMXf3Yzx+XL+Xv583ji4YIDyaDIVISsiAkoIGmjSp+bxUU/ICx38+iXzj/2DGsPZl25qIoKKqCqm0CBJpTVUaKSKf+yK4RAQz264wUKlrBa8BSIqQFh+O8PB6viIikZhZIDxarbKeHMEvvwSjv8s8fNS6bBYzVfWmUBSLIJWWQCqr41rcU1wLjEdTY3OYd20njDYKBe2WaN2mA0qKfBEVXSQM10rXeCgObVmHUcbxWDjtI9iPnoyNXipYtvVXfDtc743mb4v2uwvPPMDoQzOYaDdwX4NCHxzeeAEaH36J6YN0AQCKykpQVlGGVFqMmhd3KMQT3/uIhyK62piCxxYwxv5Jd+/eBcoHDdTH/v37YW1tDRcXF2GoiopHqxW/503UMWuoI2kGwoKfQorW+LhXDxkfJCJotWgFXQD5xSVQB4DCOJxxd4ey1SwsGN1DuEPjU1RGa6Pu6N69nq9u3WDQsq6PbHKRnJ0BqUoTtNOry8dREW7t/i9OlHbB0nXT0bYi59JpBW1NDUiLw5CSKuzxB0CaibDgpyC0xSDrHhAL440pxR8/TjKGsXH9Xp279sD0XfeFR5MtLwYbvv0ZCtNXYcuMnpWbW7TShZpKKTJzIpElo9NaYlgQQlMKYWpkix5GDXv7yyQSQVFRCZAUICM7Vxh9rWY9p2DHqs+gXxyBWz7xMJm6HmtH277htS1AyBM/FAAwMTaCds1v0jeUhz9/WovreoOwcs2IlwNbdFtBq6kK8otCkZpW/QsGFaYg7MVzAF0wyNpUGGaMMbkTEBCAdevW4ejRo8JQo2jQTyvKC8Gzu3GAajsYd5CdkKhqN4c2AFIo+7iJ+nMTfjlagokrXNGlpmFglIm/f1uFGSv/QGQd+/zIH0LFYhGi8lnaa1MY/Td27zmHh8FJOPXDbMyePRsuLi5wmbUU5wOyoCQmpGZkCncD5QXj+d04QK09jDu0FIaB9MfY+sNy/HDM/7XTQtSbpgGGztmMzZvr9/plwzp84VD3vlOJ17Zi56U4RAf8jRVzXMvOi4sLXNccQWQWUFxUICNZK0FsUBQiskrQ2tQM1efSL8Kjkz9ixlc74J9UvwpYraj8ybmEkJKRIYy+XmoADp+9ifRSTbRoro2IG0dw2DtZ2KpuKALB92MAkSF6mndBQ9bVsp+fwY4D1/H4RQQOfuMKV9eyazPLZRWuB+VBSSRBamaWcDcUpQXi6d0EQKcjjNtV739Hiffxw3fLsfV8cC2jSRlj7J+RlpaGmTNn4saNG/Wuxr0x4XPRt5FxZxOZaChQiwFz6H5yTf3VysR6baHeAOlP3klpcf60qF8r6u96gmROG5r+lJY5tqL2Y3+m2MLa+2AV+hwgly/n0O679Zj9PCuU1oxrT+3b1+9lYGRM07bdEh5NhkT6fe5AUlZpSwtPhQuDArn0+0wrMh2ygM7dvEMPHjygK1eu0JUrV+jqpd/p88GdSVG1Nc1zCxbuSBleP1NnNVCLgV/So9Tq1yD1zq80vF8nGrf1EdUypy4RET3cO4sc5q6nxw0+vf9biPOkpUMtqffcXeTr60u3b9+uPDdXDqwgi7Yq1Kz3RPorVLgjERUE09rxPQjqerTkdLQwSlSYSBvGtKe2jssoIP01ZyfSg/5v4VRaea4O/d5KImj9tJ4ElVa0yqOePbJS/Wjt+E6kpdeJ/vPrNfI58xv11VWhJl1H0KH6TrRLRJLoszSuvQah40g68exlj9EK0oJsSkpMp5La32Y1SKa90yypi9MyuuJ5m+7du/fyuvy1hyb27UBiHRP67kKccEeKOv8t6QLUetgKCs6t/otjL66lQX3Mac7h6vd7VRK6+vMEGjRvG4XnCWOMMVazitUJalq5QCg1NZUcHBzqNdGtg4MDAaD169cLQ3XWoJW1qCAfJBQowqyrJQyayz60WEkFYhUgvSQRF7Z+B4/MQVjw9SeQOf2tdlf8eDEJUScWoa1y7VWpO+dOITC4GIZmFauN1oGoCXTbm8LUtH4vExMTtNNRFx5NBnXoampDVFiMxKRUYbCKlIc7sfdMDuyc58LpI1tYW1vD3t4e9vb2sHMYA0dzI6CwCFkFhcJdERXojYRCMbp3s4ShTvVroGM7B2dvv8DxL61qXFf0pXic3H0aUDWAQfVix79Egkt7V+LgPX2sWD4DlpaW6NevX+W5sR8/BL1bNEVxQSEKiqrPQEupsYiIDoNKs+7o2b2GqVOUdfHVySjEXvwRPbRrPzvPr53DPa9E6JvVYcbh0hIUFRUCYi0006ypdCxDfjjWzBmLH68WY9jC7dgxZyB6jvgMe3Z/AZvss1i26AucD61eqapNxL3r8E7PhaGZGYzbCP4WKsbFnyZg9Lz9SK2xb5ls0Vc24ddzpRj9n9mw79sPNjY2L6/L0NGwM20P5Bciu7D6PRsZ6I1kKMOiWw+0Va/+/m7r+A3+9vLHr1NeU30tfoFTBy5BpUVntC8bcM4YY69la2sLAAgNDRWGqkhLS4OjoyPWr18vc961mlSMAq34PW9EmL29uWw6NdeWFNGMvjz4TBisIuv5SRrRHgT9jmRm2IW+Ohoko8qTQw9/30YLFiygxcu3050I2ZWE7Bgf2rx5HtkatSFNbWuaMMeFXObOozXujbmcUH2UkM+W6SSGMo3eXMto0KxnNH+gEXUfsZr8X100tVIGuS8ZQoAWfbLxpiCWTSddbUgRzWnh4edVIpK8IDr67TJauHAhrdx5mRJlDiMspbCAi7TwczvSVdCkztYjaZarC7n83zpy9/931xfNeLCLuupoksPq8zWPgky/SXMtdQlGjrTvQfVyYNz1zdRTC9Te8RequjBIEQX+tZ/mz59Pixf/SJefpVWJvqokNZh+O7CKHCwMqImiKY2Y5UIuLi60ZP91mWu1SuI8aZatCjVpOZH+jK3j6MuCMNq3YACpNtOnqSvcKbfKbiUUfHo+mSqD2vRbTPeSqldQZXmwdTSpADRymUeVNXmJiCjjHs3q3obGbrpddn7zI+nU7u9p/vwldPxe9YpYpZRHNMlCjz6YsoXCaxzIHU+/fd6foNiWJu/1FcRSafe4LgR0oG9OVR3DW5jmR3uWLKKFCxfSjwdvU0aV6KsKKdj3PM2e3Ie00JS69x9Hrq4u5PLdJroUVI8KO2Psf1JqamrlLA+yvG6tTzc3N7pw4YJwM1FZ/ycCQKmpqcJQnVUvvbypTF9cCkiEBGJoadU+Z76CggiKYgCxoVDs4YqZ47rIqPI0gb6FAUSxh7H53G2EV++iVUlZqw1MVYACaRLMJk6F6/iJcBo6BAPMaqig/CvEMOpphk7aYuSFxspYdkeKwHNbcdArAR3MHWCuKYwDQC4iEhMA5EJJWDzK8MbFgERIoATNZlWvgYJSC5haNIHPhU346ayfjFF5AKAAbd1O0CtOgXJ3IzhMmYGxo8fAyb4fzHT/zXJFPM7t/RXP0vQx+KMPax68kpmEoLxCIC4WGXnCOehK8Oz+LfhlAU11NAWrOiiiRRdT6JdcxOZTHvCOE57Yl0RqOujaXA156ZHoMNYZc8aMhpOTExx6Gsrs8J+TFoPIF4VQNe0CY50a//IqSnJDcGLBFHy+IwAfuv6EnStHQ71K5zIxOo/cgEM75kHp4XZ8Nv0beCbUpY9dDJ4+SUJhE210tOqBqrdXKf7avhZuse1ha9Gp7Pzmx+DBpSPYsmUH7kTVMJgFAFAKn2M/4VhAJrpZ28GwpoHckkyEJqcA0jSIhKc2zhMezzIANIGmZtVquFi1NSytRLh5ehM2ngmASFaHNVKETuvOaJGXAI2eFnCaMh2jRo2G00BbdNSpRyWTMfY/SUdHB+PHjwfKBw4IVVTUwsLCcPDgQTg6OlZ5dezYEXPmzMGQIUOEuyI8PBwA4ODg8Hb924TZ25soSH1GfywZQToKIECN7MdsIa8Q2RlkaeQ1+vJDkGK7AbTtzuu++UbRugGtyMJ5LcmYVqxS3IW1ZNVUlRadTxaG5EN+AK0caEAtrSeQR0zVWmJpShTd2/klfdyxCQEgS4ev6Oid5/Rq1xtJQSxd2TibLLQVCACZ9v2MNl97SIVElB//mNwWOVFzBRCgToPHbaN7YVWrSznBbuSoI6ah667JqGRWkNKNpYPJzHIYeci+jP+Y1ID7dHbtMNIRg4CWNGrxdrobWrXOkhb6kPbOtiN1BRCgRA6uG+jGi4pJvTLI7/g6GtpJhQBQ6w5jaJP7E0FlKZf+mGZOpnafk2f1olwVmT77aZCWIo3b5ScM1Sjw7ArqAAXq993ZWhcyr5D63I2+trejPov/oAwZ1boy2XR5/yTq32corb9Z4wRmlfJTAsntqxFk2QokamVKwxdvpX379pW/9tD278eROkAdR39PDyoKi2le9PVIEwLUaN6x8nn+XlEUH0pe2z4nq7ZKBCiQ7ejvyP3+iyr/j8VZ4XRh7WfUSa3sm6Wlwxe0764/lRJRZvgj2j/bntQAArTJadp+8oupWj1PuruN+mo1oUl7hBU5oQI6M6MXdftoOt2vy0lmjLFXuLm5EYAaK2dWVlaV1TFZr+XLlwt3I3rNcetDgSqGKL6FvIQHOHTkNgpELwt1ZnYT8Yl59fF2AEA5ibh67iiCmvTDtHHWqLV3WdpFjOk0G5IZW3Fwo1MtbfNxbtV4TN5OOHT/OEZ1rGtfsn+SBDfX2uPjH+Kw7LAHfhxbtgAsAJTGB+PskT8RpfjyHDbr3Bdjh/aGZnk3HkluCP7ccw7h9LJfj4Zhd0wdNRiSiDs44n4XeZXXQAFd7SfBsUfZXFcAEHJ+ORyGHcfYMxewfkQt667m+uOr0TPwt/o0nD05Dx1qLnv+YxI8L+HYg6eQvjKK1sRuKj41fznaNfHxFRy7HADJK/dg54HDMcyyI4BUeB06hQepL6fN0O7wIZxG26JFxSFLfLHIagp8u3yJwydcUT6rXQ2kuLNjGj796gXWXDmLeX1qvscrSVOwd5otZl1vh4NnT2Jar7p8s8pDUnQ6JFrt0Eb2DV8uDdFhhWhm3FZQKasqL94HJ367hHSVskpTaWkp8vPzQUQQi8VQVlaGWCyG0QefwnGACVQBoCgR1z084Beegx6DJ8C+e9UR3gXhATjlfhlJii9vkOZdB2KCo2Xl2p5F6U9xep8H4l9po23SC9M+7Y/MZ3/jiIcfSiuvmTKsnCbjo04VdU+C94E5GD7DE/NuX8bSfm0rj1FNhiemO85GuPlSuO+ZghrGQTPGWK169eoFY2Pjt17H81XOzs4ICwvDo0ePhKH6EWZv8ibm3FfUzdyWvjxYtQ9WNQVhtGa4Lon6LaeIXGFQfuRGeND07hpkMuEXiq6x41VjKSL3JVYkNptA54JqHypX/OQwDbE1oL7fXBGG3lvZ97eRvY0Fjfj+Vu1VR0km7ZluQLByobvxtbYkIqLcOz+RqZYq2c77g2oYnMtqI8mk3TM6k6LZf+h2bIkwWkXu3c1ka92Fhv9cS39Qxhirhb+/PwGo10jP2oSFhTXY8Rquz1ojiX0RDFIFOnYxEIaqyInxh8+9dPTv/wFayWNRrZy6wWDMmjsa2Zd+xv9tulNtIftGI8lG2LMg6HXqAMPWtfU9I4T4+yEqqAAf9637aJd3XXJ4CLKKMmDU1URG/8kykqxA+HglwLqXJYx0a2sJoCQYu37aguQuY7Bq3hjUMDiX1aYwHWGBYdA1M0a7FrJ6BAKABMH+T5AeS+hv21UYZIyxOjE3N4ebmxtmzpwpDL2R5cuXw83NrV4jR2WR748PaQJehORCKumGD81VgYJ8QFLWcbwkyhP7Dh3BzdCylaHzs+IQl6yB9vptoQYA2RnISitCwVs/5G1oirCZ+TP2fWWFO5sXYeHemzIGGzSskoQH8H2qBjPjHjDUBJBbMWusFOE3f8P6QxcRW776eXZqFrLTO8DAsHzOjrh4xKbUpQP7uyobEeHJSE42wcd9WwHFhUBJeU/4lAC4HdmPs/5l062UFMQhNkwDerrt0VwEoCAHRRnFyBGMZygJuYk9cydiXUY/rN72E+yNXz+wgFWVG30PfoGa+KBrd+grv3rPluKxx078cPgq0orLfs5IyEVBjiEMOpR9U5PExiM+4x/7KsQYe084Oztj8eLFWLFihTBUL3PmzMGIESPg7OwsDL0R+U7WKANx8VkoLtFDRsRNHNhzAPdCyjIKSeQtHD9zDleCy34WK2uiXbNSPH9yC89DQ3F84/c4cT1cPmc8V9TBJ0sPYt+aAQjb64pJM+bjtF96o/6tpTkpyMjVgppYHQH3T+KA23UkFACAAsKv78O2056Iyin7WSRWRHO1KDx6cAfhT2/h7JEz8Iooz+TeSzmIjopDYWk7FMbfw+E9O3DZt2xeHKQG4OKfJ+D2ILHsZ0UNtNMrRWz4HQSERsNj5wa4Xw1FUcU7SZqFxwe/x5J5q7A/+GMc2LkXc61lr+bBZCvOTkFWUUs0U1PF3VtH8Nupe0iXAIAEkbcPYcvp+0gqBAARlJRF0FAKgfcDb4T6XsKFE1fxqE4jZBljrCpnZ2eMGzcOGzZsEIbqZMOGDXBxcWmwRA0AFFetWrVKuFFuiHRg2FqK4JCH8Huuhf7Oo9C7Y3MoKSqgINILp+6nw3zwcNjoK0O1hT70VTPh6xcArztpMB37OYZ/1A5NlWt7fPIvUlCBUU97jHKyh3aaL2LFRvigs26jZc9inY7Q1QjH3ft+yFa2wbAJA6CvoQSRggJCbhyBV44ZJo7qB11lQM9ADxrFKbjrdRu+4S1hPXkk7MzKzvv7SQNG7TSQmOSDO3cksBk7AQO666KJWASkPsGpq8+g++EoOJo2g1i9Ldo3z0fAI3943oqE7sCpGGpvBB1lMRQAoDACl88GovnwOfjvms9gocsVtTel0qoTmikF4v7dJyjVGQSnsTbQVVaCSEGKR2f3wAe9MM3JBtpKIugbtoIoMw5enrfwNMkAtpOGop9xU4hF7+s9yxhrTHp6eujTp49wc5306dMHenp6ws1vpUFGg/7zCnB3w1xsjLTBll9moT1PpfTGpAXPsWn6YkTa/YiNMy0qR/ExAJAi6Lcl+PqSFtbs+hYW2o2VSrP6KM24j1VTV6Jo4lb8MKFLzXPuMcbYe+Sd/PQpDrqAe2p9sGzpVE7U3lLotROI77MQ30/jRK2amNt4QJ0wffkiTtTkSMDVMygdvhKrxnOixhj73/COVtYYY4wxxv43cLmAMcYYY0yOcbLGGGOMMSbHOFljjDHGGJNjnKwxxhhjjMkxTtYYY4wxxuQYJ2uMMcYYY3KMkzXGGGOMMTnGyRpjjDHGmBzjZI0xxhhjTI5xssYYY4wxJsc4WWOMMcYYk2OcrDHGGGOMyTFO1hhjjDHG5Bgna4wxxhhjcuz/AQVSy9/5liwuAAAAAElFTkSuQmCC\" style=\"width: 470px; height: 46.3166px;\" width=\"470\" height=\"46.3166\"\u003e\u003c/p\u003e\u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{A{I}_{i,t}}\\)\u003c/span\u003e\u003c/span\u003e is the industry\u0026rsquo;s annual artificial intelligence mean after excluding firm i\u0026rsquo;s own data\u0026zwnj;. This instrumental variable considers exogenous shocks at the industry level, the intensity of firms\u0026rsquo; use of artificial intelligence technology within the industry, and the lagged level of artificial intelligence inputs to meet the criteria of relevance and exogeneity.\u003c/p\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e6\u003c/span\u003e presents the regression results estimated using two-stage least squares (2SLS). Columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) display the regression outcomes for the two instrumental variables in the first stage, with F-values exceeding 10, satisfying the criterion for instrumental variable correlation. The Cragg-Donald Wald F-statistics are above the Stock-Yogo weak instrumental variable test\u0026rsquo;s 10% critical value of 16.38, passing the test for weak instrumental variables. The Kleibergen-Paap rk LM statistics are significant at the 1% level, rejecting the null hypothesis of \u0026ldquo;insufficient identification of instrumental variables,\u0026rdquo; thus indicating no under-identification issue. Columns (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) and (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) detail the second-stage regression results for the two instrumental variables, where the coefficients for dAI are significantly positive (0.046 and 0.065, respectively). This suggests that the baseline regression results remain robust even after addressing the endogeneity problem, thereby reaffirming that artificial intelligence information disclosure encourages institutional investors to increase their shareholdings.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eInstrumental variable approach\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003edIO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003edIO\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.046*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.065**\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(1.865)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(2.429)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eL.dAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.248***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(-11.498)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBartik_iv\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.051***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(4.393)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e27908\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e27908\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e27858\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e27858\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj.R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.109\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.108\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eF statistic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e132.21***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e19.30***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCragg-Donald Wald Fstatistic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1741.004***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1694.386***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKleibergen-Paap rk LM statistic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e59.003***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e13.837***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"5\"\u003eNote: This table reports the results of two-stage least squares regressions, with columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) being the lagged one period for the first instrumental variable, dAI; and columns (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) and (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) being the second instrumental variable, the Barkit instrumental variable. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec33\" class=\"Section3\"\u003e\u003ch2\u003e5.3.2 The two-stage Heckman\u003c/h2\u003e\u003cp\u003eFirms\u0026rsquo; decisions on whether to adopt artificial intelligence technology are typically not random and may be influenced by unobserved factors such as management preferences (Liu et al., 2025b) and technological capabilities. Companies that are more inclined to implement artificial intelligence may be more likely to disclose AI-related information in their annual reports, whereas those less focused on AI may provide less disclosure. This disparity could result in sample self-selection bias. Consequently, this paper employs the Heckman two-stage method to assess whether the sample exhibits a self-selection bias issue(Heckman, 1979).\u003c/p\u003e\u003cp\u003eIn the first stage, a dummy variable is constructed based on the mean value of artificial intelligence information disclosure for each industry-year, with values higher than the mean assigned to 1 and 0 otherwise (Chen et al., 2025). Additionally, the explanatory variable, lagged one period (L.dAI), is included as an exogenous variable. And we use a probit model to test whether the one-lagged control variables affect artificial intelligence (Jin et al., 2024). Next, we calculate the adjustment term, the inverse Mills ratio. In the second stage, the calculated inverse Mills ratio (IMR) is incorporated into the baseline regression Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) as a control variable to correct for estimation bias resulting from sample selection bias.\u003c/p\u003e\u003cp\u003eAs indicated in columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, the impact of the IMR is negligible, and the coefficient for dAI remains significantly positive at the 1% level. This indicates that, after accounting for sample selection bias, the impact of artificial intelligence information disclosure on institutional investor ownership remains robust.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab7\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eEndogeneity test\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePhase I\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003ePhase II\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePSM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eIndustry fixed effects\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eProvince fixed effects\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.014***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.018*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.013**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.013**\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2.653)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(1.896)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(2.398)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(2.448)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eL.dAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.080***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(-8.399)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eimr\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.033\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.278)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInd FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProvince FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e_cons\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-1.052***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.596***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.964***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.254***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e3.019***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(-4.459)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(7.237)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(3.197)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(11.222)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(10.782)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e28158\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e27908\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e30964\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e32256\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj.R2\u003c/p\u003e\u003cp\u003e/ PseudoR2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.025\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.061\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.334\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.057\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.057\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"6\"\u003eNote: This table reports three endogeneity tests, columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) are Heckman two-stage; column (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) is PSM; column (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) adds industry fixed effects, and column (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) adds province fixed effects. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, Ind FE denotes firm fixed effects, Province FE denotes year fixed effects. and standard errors are clustered at the firm level. The values in the parentheses in the first column are z-values. The values in the other four bracketed columns are t-values. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec34\" class=\"Section3\"\u003e\u003ch2\u003e5.3.3 Propensity score matching\u003c/h2\u003e\u003cp\u003eReferring to Li et al.(2023), we define the dummy variable (AI_dummy) by taking the mean value of artificial intelligence applications as the basis for division. Those applications with values higher than the mean are categorized into the treatment group (AI_dummy\u0026thinsp;=\u0026thinsp;1), while those with values lower than the mean are categorized into the control group (AI_dummy\u0026thinsp;=\u0026thinsp;0). We utilize Size, Lev, Roa, Age, Independence, Top1, Return, and R\u0026amp;D background of the executive team as matching variables for one-to-one nearest-neighbor propensity score matching with a caliper of 0.05. The balance test results indicate that the standardized deviations of the covariates post-matching are below 5%, and the t-test results support the initial hypothesis that the coefficients of the treatment and control groups are not significantly different. This suggests that the differences in characteristics between the experimental and control groups have been largely eliminated, and the matching effect is satisfactory. We rerun the regression using the matched samples, and the results are reported in column (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e. We find that the coefficient for dAI is 0.018, which is significantly positive at the 10% level, and we continue to observe a positive correlation between artificial intelligence information disclosure and institutional investor shareholding.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec35\" class=\"Section3\"\u003e\u003ch2\u003e5.3.4 Industry-fixed effect and Province-fixed effect\u003c/h2\u003e\u003cp\u003eTo address the issue of omitted variables, this paper incorporates industry fixed effects and province fixed effects into the regression analysis. The results are presented in columns (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) and (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, indicating that the coefficients for dAI are significantly positive at the 5% level, corroborating the findings of the benchmark regression.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec36\" class=\"Section2\"\u003e\u003ch2\u003e5.4 Heterogeneity test\u003c/h2\u003e\u003cdiv id=\"Sec37\" class=\"Section3\"\u003e\u003ch2\u003e5.4.1 Period heterogeneity\u003c/h2\u003e\u003cp\u003eThe \u0026ldquo;Industry 4.0 policy\u0026rdquo; was initially proposed in Germany in 2011, with the goal of fostering the digital transformation of the manufacturing sector. Subsequently, the world has embarked on the fourth industrial revolution, and governments have expedited the advancement of artificial intelligence technology. Consequently, this paper uses 2011 as a pivotal year, dividing the sample into pre-2011 and post-2011 periods for heterogeneity analysis (Tao et al., 2024).\u003c/p\u003e\u003cp\u003eAs evident from columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, the coefficients for dAI are insignificant prior to 2011 but significantly positive thereafter. This indicates that the impact of artificial intelligence on institutional investor ownership has become more pronounced post-2011. The potential reason for this shift could be that following 2011, China hastened the development of artificial intelligence, influenced by the fourth industrial revolution, to drive changes in the industrial structure.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab8\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 8\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eHeterogeneity analysis\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"9\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c9\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2000\u0026ndash;2010\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2011\u0026ndash;2020\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-manufacturing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eManufacturing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eHigh Competition\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eLow Competition\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eLow R\u0026amp;D background\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003eHigh R\u0026amp;D background\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.135\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.012**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.018\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.015***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.022**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.003\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e0.019***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(1.396)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2.237)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(-1.051)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(2.723)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(2.397)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e(0.395)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e(0.076)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e(3.260)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e_cons\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6.539**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.525***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e3.083***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.425***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e3.039***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e3.534***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e3.197***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e3.349***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(2.319)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(7.440)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(5.995)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(8.871)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e(6.713)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e(7.812)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e(7.450)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e(5.327)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4007\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e28134\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e9827\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e22400\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e15573\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e16261\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e17467\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e13322\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj.R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.025\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.043\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.034\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.067\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.049\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.053\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.032\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e0.064\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ep-value\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003e0.000***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e0.004***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e0.072*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003e0.060*\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"9\"\u003eNote: This table reports the results of the analysis of heterogeneity, with columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) based on the 2011 subgroups; columns (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) and (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) for the subgroup regressions of manufacturing firms and non-manufacturing firms; columns (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) and (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e) for the subgroups based on median HHI; and columns (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e) and (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e) for the subgroups based on whether the chairman of the board of directors has an R\u0026amp;D background. Tests for differences between groups are Fisher\u0026rsquo;s combined tests, all with 500 samples. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec38\" class=\"Section3\"\u003e\u003ch2\u003e5.4.2 Industry heterogeneity\u003c/h2\u003e\u003cp\u003eManufacturing serves as the cornerstone of the real economy and is a significant vehicle for the innovation strategy of artificial intelligence, playing a crucial role in the process of high-quality economic development. Given this, we categorize the sample into manufacturing and non-manufacturing industries (Wu et al., 2025). The results of the grouped regressions are presented in columns (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) and (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, indicating that the impact of dAI on dIO is more pronounced in manufacturing firms. Conversely, no significant effect is noted in the non-manufacturing samples. The potential reason for this discrepancy is that manufacturing firms utilize AI technology to substitute for large equipment, thereby reducing fixed asset costs and enhancing productivity, which in turn encourages institutional investors to boost their shareholdings. In contrast, the non-manufacturing sector has a higher proportion of employees in sales and administrative roles, which involve tasks that are less conventional, such as social interactions-tasks that artificial intelligence currently cannot fully replicate (Wu et al., 2024).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec39\" class=\"Section3\"\u003e\u003ch2\u003e5.4.3 Competitive heterogeneity\u003c/h2\u003e\u003cp\u003eReferring to Haushalter et al. (2007) and Ren et al. (2025), we measure the degree of competition in the industry using the Herfindahl index, or HHI, which is the cumulative sum of the squares of the ratios of each firm\u0026rsquo;s operating revenues to the industry\u0026rsquo;s total operating revenues. We stratify the sample into high and low competition groups based on the median of the HHI (Yang et al., 2025), where smaller values of the HHI indicate more intense competition.\u003c/p\u003e\u003cp\u003eWe observe that artificial intelligence information disclosure plays a more significant role in determining institutional investor ownership within firms operating in highly competitive industries (refer to column(\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) in Table\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). Conversely, there is no notable effect in the low competition group. The potential reason for this is that in highly competitive sectors, firms with increased artificial intelligence information disclosures tend to decrease information asymmetry, thereby enabling external institutional investors to access more information to inform their investment decisions.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec40\" class=\"Section3\"\u003e\u003ch2\u003e5.4.4 Heterogeneity of board R\u0026amp;D background\u003c/h2\u003e\u003cp\u003eThe board of directors is the key organization responsible for R\u0026amp;D expenditure decisions. Directors with specific technology backgrounds can provide unique insights to guide corporate R\u0026amp;D investments. They communicate and collaborate with the CEO to ensure that the company\u0026rsquo;s R\u0026amp;D strategy aligns with the overall corporate strategy and play a crucial role in promoting the development of corporate artificial intelligence. Consequently, we follow Li et al. (2025) to calculate the ratio of directors with R\u0026amp;D backgrounds, grouping the observations according to the median of this indicator.\u003c/p\u003e\u003cp\u003eAs evident from columns (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e) and (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab9\" class=\"InternalRef\"\u003e9\u003c/span\u003e, the coefficient of dAI on dIO is 0.019 and significant at the 1% level in the sample with a high R\u0026amp;D background. However, it is not significant in the sample with a low R\u0026amp;D background. This suggests that the impact of artificial intelligence information disclosure on institutional investor shareholdings is greater in the sample with a high R\u0026amp;D background. A possible reason for this is that directors with a technological background are more inclined to allocate more resources to strengthen their firm\u0026rsquo;s technological capabilities (Ginesti et al., 2025) and to promote the development of artificial intelligence technology.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab9\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 9\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eInstitutional investor site visits\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003edSV1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003edSV2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003edSV3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eSV4\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.022***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.017**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.016**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.033***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(3.294)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2.391)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(2.310)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(5.464)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e_cons\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.749**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.294\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.239\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.233\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(2.022)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(0.760)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(0.626)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(0.562)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25412\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e25412\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e25412\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e26925\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj.R2\u003c/p\u003e\u003cp\u003e/Pseudo R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.011\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.028\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.032\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.0682\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"5\"\u003eNote: This table reports the results of the mediation mechanism for institutional investor fieldwork. dSV1, dSV2, and dSV3 are differenced and standardized variables for the number of institutional surveys, the number of institutional surveys, and the number of people researched. SV4 is a dummy variable that equals 1 if an institutional investor site visit occurs to the enterprise in the current year, and 0 otherwise. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, and standard errors are clustered at the firm level. The values in parentheses in columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e), (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), and (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) are t-values, and the values in parentheses in column (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) are z-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec41\" class=\"Section2\"\u003e\u003ch2\u003e5.5 Mechanism analysis\u003c/h2\u003e\u003cp\u003eThe previous theoretical analysis indicates that the disclosure of artificial intelligence information can encourage institutional investor shareholdings by prompting institutional investors to conduct fieldwork, enhancing information transparency, and fostering innovation. Consequently, this paper concentrates on the impact of artificial intelligence information disclosure on these three avenues and constructs the following model to test the mechanism of action:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{dM}_{i,t}={\\varphi\\:}_{0}+{\\varphi\\:}_{1}{dAI}_{i,t}+{\\varphi\\:}_{2}Control{s}_{i,t}+{\\lambda\\:}_{i}+{\\mu\\:}_{t}+{\\epsilon\\:}_{i,t}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere subscripts i and t denote enterprises and years, respectively; M denotes mechanism variables; and other variable settings are consistent with the benchmark regression Eq.\u0026nbsp;(\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cdiv id=\"Sec42\" class=\"Section3\"\u003e\u003ch2\u003e5.5.1 Institutional investor site visits\u003c/h2\u003e\u003cp\u003eWe follow the methodologies of Jiang et al. (2018) and Cao et al. (2024) to select the number of institutional investor visits (SV1) as the first proxy variable for fieldwork; refer to Zhao et al. (2023) to select the number of institutions visited (SV2) as the second proxy variable for fieldwork; following Xu et al. (2025), the number of institutional investors\u0026rsquo; research (SV3) is selected as the third proxy variable for fieldwork; regarding (Lai et al., 2022), the dummy variable for institutional investors\u0026rsquo; fieldwork (SV4) is selected as the fourth proxy variable for fieldwork, and takes the value of 1 if the company has institutional investors\u0026rsquo; fieldwork in that year, and 0 otherwise.\u003c/p\u003e\u003cp\u003eAs indicated in Table\u0026nbsp;\u003cspan refid=\"Tab9\" class=\"InternalRef\"\u003e9\u003c/span\u003e, corporate artificial intelligence information disclosure influences the indicators across various dimensions of institutional investor fieldwork. Initially, the disclosure of artificial intelligence information notably boosts the quantity of institutional research. This suggests that following enterprises\u0026rsquo; release of artificial intelligence data, institutional investors are more inclined to communicate and interact with these companies, thereby gaining a better understanding of their business operations and the development of artificial intelligence technologies. Secondly, the number of institutional research studies sees a significant rise, implying that the disclosure of artificial intelligence attracts a broader range of institutional investors with diverse backgrounds and investment strategies to conduct their own investigations. Thirdly, there is a notable increase in the number of researchers, signifying that institutions are eager to allocate more personnel to participate in the research, which also substantially enhances the likelihood of field research taking place. In summary, artificial intelligence information disclosure fosters various aspects of institutional investors\u0026rsquo; fieldwork activities. Following site visits, institutional investors will then determine whether to augment or reduce their stock holdings.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec43\" class=\"Section3\"\u003e\u003ch2\u003e5.5.2 Information transparency\u003c/h2\u003e\u003cp\u003eFirstly, concerning Clarkson et al. (2008), we construct the Janis-Fadner coefficient, an indicator of media attention, as the first proxy variable of information transparency, as shown in Eq.\u0026nbsp;(\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e4\u003c/span\u003e):\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:J-F=\\left\\{\\begin{array}{c}\\frac{{e}^{2}-ec}{{t}^{2}}\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:if\\:e\u0026gt;c\\\\\\:\\frac{{ec-c}^{2}}{{t}^{2}}\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:if\\:e\u0026lt;c\\\\\\:0\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:if\\:e=c\\end{array}\\right.$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere e is the number of positive media reports, c is the number of negative media reports, and t is the sum of the number of positive reports and the number of negative reports. The value of the J-F coefficient ranges from \u0026minus;\u0026thinsp;1 to 1. The more positive reports about the enterprise, the closer the J-F coefficient is to 1, and the less pressure the enterprise faces from media monitoring. When there are more negative reports about the enterprise, the J-F coefficient is close to -1, indicating that the enterprise faces more media monitoring pressure.\u003c/p\u003e\u003cp\u003eSecondly, following Weng et al. (2024), we select the number of analyst coverage, the number of analysts tracking a company in a year, as the second proxy variable of information transparency.\u003c/p\u003e\u003cp\u003eThe regression results in columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab10\" class=\"InternalRef\"\u003e10\u003c/span\u003e indicate that firm-level artificial intelligence information disclosure significantly enhances media monitoring and analyst coverage. Additionally, the results suggest that artificial intelligence information disclosure fosters institutional investor ownership by enhancing information transparency.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab10\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 10\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eInformation Transparency\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003edmedia\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003edanalyst\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.015**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.022**\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(2.094)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(2.542)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e_cons\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.618**\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.647***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(2.401)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(3.364)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e29967\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e19012\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj.R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.070\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.092\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"3\"\u003eNote: This table reports the results of the mediation mechanism of information transparency. dmedia and danalyst are differenced and standardized variables for the J-F coefficient and the number of analyst coverage. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and the values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe possible reasons are that the media enables institutional investors to make objective judgments about the development of corporate artificial intelligence through reports on AI investments and AI patents. With their extensive knowledge, analysts can provide institutional investors with professional reports, which are crucial for them to decide whether to increase their stock holdings (Li et al., 2024).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec44\" class=\"Section3\"\u003e\u003ch2\u003e5.5.3 Innovation\u003c/h2\u003e\u003cp\u003eIn this paper, we reference existing research (Yuan et al., 2018) and calculate the sum of invention patents, utility model patents, and design patents to measure the innovation level of enterprises (patents). Table\u0026nbsp;\u003cspan refid=\"Tab11\" class=\"InternalRef\"\u003e11\u003c/span\u003e presents the results of the mediation test for the innovation level of enterprises. We report that dAI is significant at the 10% level, indicating that changes in artificial intelligence information disclosure can promote the innovation level of enterprises. Innovation, as a key indicator of the long-term development of enterprises, has an important impact on the stock selection decision of institutional investors. When an enterprise\u0026rsquo;s innovation ability is enhanced, it not only significantly increases the success rate of enterprise R\u0026amp;D but also sends a positive signal to institutional investors about the long-term viability of the enterprise, which encourages them to increase their stock holdings.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab11\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 11\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eInnovation\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003edpatent\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.031*\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(1.658)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e_cons\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.252\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(1.461)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e31352\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj. R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-0.102\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"2\"\u003eNote: This table reports the results of the innovation mediation mechanism. dpatent is the difference-in-differences and standardization after summing invention patents, utility model patents, and design patents (patent). Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and the values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"6. Further analysis","content":"\u003cp\u003eAlthough institutional investors share certain commonalities, they must not be viewed as a homogeneous group (Yan et al., 2009). Though institutional investors emphasize the long-term value of firms, different types of institutional investors have varying levels of concern for the long-term value of firms (Cox et al.,2011). To investigate the impact of artificial intelligence information disclosure on institutional investors, this paper categorizes them into short-term and long-term institutional investors, as well as stress-sensitive and stress-resistant institutional investors.\u003c/p\u003e\u003cdiv id=\"Sec46\" class=\"Section2\"\u003e\u003ch2\u003e6.1. Short- and long-term institutional investors\u003c/h2\u003e\u003cp\u003eReferring to Yan et al.(2009), we construct the indicators of short-term institutional investors and long-term institutional investors according to the following steps.\u003c/p\u003e\u003cp\u003eFirstly, we calculate the cumulative total stock assets bought and sold by each institutional investor in each half-year from 2007 to 2023. For a given institutional investor k, the cumulative total assets of stocks bought and sold per half-year are computed as follows, respectively:\u003cdiv id=\"Equ4\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}C{R}_{-}bu{y}_{k,t}=\\sum\\:_{i=1}^{N}\\:\\left|{S}_{k,i,t}{P}_{i,t}-{S}_{k,i,t-1}{P}_{i,t-1}-{S}_{k,i,t-1}{\\Delta\\:}{P}_{i,t}\\left|\\right({S}_{k,i,t}⩾{S}_{k,i,t-1})\\right.\\end{array}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e5\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equ5\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ5\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}C{R}_{-}{sell}_{k,t}=\\sum\\:_{i=1}^{N}\\:\\left|{S}_{k,i,t}{P}_{i,t}-{S}_{k,i,t-1}{P}_{i,t-1}-{S}_{k,i,t-1}{\\Delta\\:}{P}_{i,t}\\left|\\right({S}_{k,i,t}\u0026lt;{S}_{k,i,t-1})\\right.\\end{array}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e6\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{i,t}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{i,t-1}\\)\u003c/span\u003e\u003c/span\u003e are the stock prices of the listed company \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e in period t and t-1, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{k,i,t}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{k,i,t-1}\\)\u003c/span\u003e\u003c/span\u003e denote the number of shares held by institutional investor k in listed company\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e in period t and t-1, respectively. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C{R}_{-}bu{y}_{k,t}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C{R}_{-}{sell}_{k,t}\\)\u003c/span\u003e\u003c/span\u003e denote the cumulative market value of shares bought and sold by institutional investor k in period t, respectively.\u003c/p\u003e\u003cp\u003eSecondly, the turnover rate of institutional investor k in period t is defined as:\u003cdiv id=\"Equ6\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ6\" name=\"EquationSource\"\u003e\n$$\\:C{R}_{k,t}=\\frac{2min(C{R}_{-}bu{y}_{k,t},C{R}_{-}sel{l}_{k,t})}{\\sum\\:_{i=1}^{{N}_{k}}\\:({S}_{k,i,t}{P}_{i,t}+{S}_{k,i,t}{P}_{i,t-1})}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e7\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThirdly, the average turnover rate of institutional investor k is calculated, whose average turnover rate is defined as the average turnover rate over the last two years (4 semiannual periods):\u003cdiv id=\"Equ7\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ7\" name=\"EquationSource\"\u003e\n$$\\:AV{G}_{-}C{R}_{k,t}=\\frac{1}{4}\\sum\\:_{j=0}^{3}\\:C{R}_{k,t-j}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e8\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eFourthly, institutional investors are categorized based on their average turnover rate. To avoid the influence of outliers, those holding investments for two or more years between 2007 and 2023 are divided into three groups following a sorting by their average turnover rate. The group exhibiting the highest average turnover rate is classified as short-term institutional investors, while the group with the lowest turnover rate is classified as long-term institutional investors.\u003c/p\u003e\u003cp\u003eAccordingly, the percentage of ownership by short-term institutional investors (SIO) and the percentage of ownership by long-term institutional investors (LIO) in each stock are calculated separately.\u003c/p\u003e\u003cp\u003eFrom columns (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab12\" class=\"InternalRef\"\u003e12\u003c/span\u003e, the coefficient of dAI is significantly positive for the regression result of dLIO, but insignificant for dSIO. This suggests that long-term institutional investors are more attentive to the long-term value investment of companies and focus on the long-term value that artificial intelligence technology brings (Cremers et al., 2020).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab12\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 12\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eFurther analysis\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003edSIO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003edLIO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003edNIS\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003edIS\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003edAI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.007\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.010*\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.004\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.010**\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(1.207)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(1.855)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(-0.879)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(2.068)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControls\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear FE\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e_cons\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3.678***\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.041\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.458\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.524***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e(11.751)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e(-0.145)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e(1.501)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e(11.447)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e31235\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31235\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e31235\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e31235\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eadj.R2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.042\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.063\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-0.081\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.109\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"5\"\u003eNote: This table reports the results of the further analysis. The short-term institutional investors ownership (SIO) is first differenced and then standardized to obtain dSIO. The long-term institutional investors ownership (LIO) is first differenced and then standardized to obtain dLIO. The pressure-sensitive institutional investors ownership (NIS) is first differenced and then standardized to obtain dNIS. The pressure-resistant institutional investors ownership (IS) is first differenced and then standardized to obtain dIS. Code FE denotes firm fixed effects, Year FE denotes year fixed effects, standard errors are clustered at the firm level, and the values in parentheses are t-values. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec47\" class=\"Section2\"\u003e\u003ch2\u003e6.2 Stress-sensitive, stress-resistant institutional investors\u003c/h2\u003e\u003cp\u003eFurthermore, following Brickley et al. (1988) and Liu et al. (2024), institutional investors are classified into pressure-sensitive (NIS) and pressure-resistant (IS) categories. Pressure-sensitive institutional investor ownership encompasses the combined shareholding percentages of insurance companies, trust companies, comprehensive brokerage firms, and corporate annuities that engage directly with the company. Their investment decisions are susceptible to management influence, and their inclination to participate in corporate governance is typically low. In contrast, pressure-resistant institutional investor ownership includes the combined shareholding percentages of Securities Investment funds, social security funds, and QFII. These investors have no direct business relations with the company and often base their voting decisions on independent analysis, which tends to enhance the corporate governance standards. This category of institutional investor is primarily focused on the long-term value of the investment.\u003c/p\u003e\u003cp\u003eThe results in columns (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) and (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) of Table\u0026nbsp;\u003cspan refid=\"Tab12\" class=\"InternalRef\"\u003e12\u003c/span\u003e suggest that artificial intelligence information disclosure attracts stress-resistant institutional investors, rather than stress-sensitive ones, to increase their stock holdings. On the one hand, artificial intelligence information disclosure can reduce the degree of information asymmetry, providing stress-resistant institutional investors with more information about artificial intelligence. On the other hand, it offers signals about the long-term development of firms, indicating that these firms possess unique resource advantages and robust R\u0026amp;D capabilities. This attracts investors to increase their stock holdings, allowing them to benefit from the future development of artificial intelligence.\u003c/p\u003e\u003c/div\u003e"},{"header":"7. Conclusion","content":"\u003cp\u003eAs a pivotal force in the new wave of technological revolution and industrial transformation, artificial intelligence significantly contributes to fostering corporate innovation, enhancing productivity, and propelling economic growth. However, the impact of artificial intelligence information disclosure on institutional investor shareholding remains inconclusive. This paper examines how changes in corporate artificial intelligence information disclosure influence institutional investor shareholdings. Based on the dataset of Chinese listed companies from 2007 to 2023, we find that the disclosure of artificial intelligence information by corporations encourages institutional investors to augment their equity stakes. This conclusion is supported by five robustness tests and a battery of endogeneity assessments. The robustness tests encompassed substituting dependent variables, substituting independent variables, elevating the clustering level to the industry level, excluding municipal samples, and altering the regression approach. To mitigate endogeneity issues, we employed instrumental variables in a two-stage least squares regression. To counteract sample self-selection bias, we conducted a Heckman two-stage test. Additionally, we utilized Propensity Score Matching (PSM) to address selection bias in the sample. We also considered the potential problem of omitted variables by incorporating industry and province fixed effects, respectively.\u003c/p\u003e\u003cp\u003eNext, we perform heterogeneity analyses, examining time heterogeneity, industry heterogeneity, competition heterogeneity, and director technology background heterogeneity. We find that the contribution of changes in artificial intelligence information disclosure to changes in institutional investor ownership is more significant in post-2011 samples, within the manufacturing industry, in environments with a high degree of competition, and among companies with a high proportion of directors with technology backgrounds. Following this, we conduct the mechanism analysis to investigate whether artificial intelligence information disclosure affects institutional investor ownership by promoting site visits from institutional investors, enhancing information transparency, and improving innovation. This serves to test the three proposed hypotheses. Finally, we further analyze the impact of artificial intelligence information disclosure on the shareholding ratios of different types of institutional investors. Specifically, we classify institutional investors into long-term and short-term investors, as well as stress-resistant and stress-sensitive investors.\u003c/p\u003e\u003cp\u003eAs evidenced by the literature, the decision of institutional investors to hold shares in companies for investment purposes is, to a certain extent, dependent on the development prospects, operational capabilities, and financial status of those enterprises. Furthermore, their emphasis on advanced technology during the development process is underscored by the disclosure of artificial intelligence. This paper, therefore, contributes to the existing literature on institutional investors\u0026rsquo; stockholding and addresses the gap concerning whether artificial intelligence influences institutional stockholding.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding:\u003c/h2\u003e\u003cp\u003eThis work was partially supported by the National Social Science Fund of China [No. 20BJY260], and BTBU Digital Business Platform Project by BMEC.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eFeng Zhao: Conceptualization, Writing - Review \u0026amp; Editing, and Funding acquisition;Dandan Nan: Data curation, Methodology, Software, and Writing;Yao Wu: Methodology, Validation, Writing - Review \u0026amp; Editing and Supervision\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData are available upon request by email.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAcemoglu D, Autor D, Hazell J, et al., 2022. Artificial intelligence and jobs: evidence from online vacancies[J]. Journal of Labor Economics, 40: S293-S340.\u003c/li\u003e\n\u003cli\u003eAggarwal R, Erel I, Ferreira M, et al., 2011. Does governance travel around the world? Evidence from institutional investors[J]. Journal of Financial Economics, 100(1): 154-181.\u003c/li\u003e\n\u003cli\u003eAghion P, Van Reenen J, Zingales L, 2013. Innovation and institutional ownership[J]. American Economic Review, 103(1): 277-304.\u003c/li\u003e\n\u003cli\u003eAgnese P, Arduino F R, Bruno E, et al., 2024. On the road to sustainability: the role of board characteristics in driving ESG performance in Africa[J]. Socio-Economic Planning Sciences, 95: 101994.\u003c/li\u003e\n\u003cli\u003eAhmed I, Jeon G, Piccialli F, 2022. From artificial intelligence to explainable artificial intelligence in industry 4.0: a survey on what, how, and where[J]. IEEE Transactions on Industrial Informatics, 18(8): 5031-5042.\u003c/li\u003e\n\u003cli\u003eAleem M, Sufyan M, Ameer I, et al., 2023. Remote work and the COVID-19 pandemic: an artificial intelligence-based topic modeling and a future agenda[J]. Journal of Business Research, 154: 113303.\u003c/li\u003e\n\u003cli\u003eAlonso C, Berg A, Kothari S, et al., 2022. Will the AI revolution cause a great divergence?[J]. Journal of Monetary Economics, 127: 18-37.\u003c/li\u003e\n\u003cli\u003eAnnan R, Qingge L, 2025. Artificial intelligence in COVID-19 research: a comprehensive survey of innovations, challenges, and future directions[J]. Computer Science Review, 57: 100751.\u003c/li\u003e\n\u003cli\u003eAppel I R, Gormley T A, Keim D B, 2016. Passive investors, not passive owners[J]. Journal of Financial Economics, 121(1): 111-141.\u003c/li\u003e\n\u003cli\u003eArmstrong C S, Core J E, Taylor D J, et al., 2011. When does information asymmetry affect the cost of capital?[J]. Journal of Accounting Research, 49(1): 1-40.\u003c/li\u003e\n\u003cli\u003eBabina T, Fedyk A, He A, et al., 2024. Artificial intelligence, firm growth, and product innovation[J]. Journal of Financial Economics, 151: 103745.\u003c/li\u003e\n\u003cli\u003eBag S, Pretorius J H C, Gupta S, et al., 2021. Role of institutional pressures and resources in the adoption of big data analytics powered artificial intelligence, sustainable manufacturing practices and circular economy capabilities[J]. Technological Forecasting and Social Change, 163: 120420.\u003c/li\u003e\n\u003cli\u003eBahoo S, Cucculelli M, Qamar D, 2023. Artificial intelligence and corporate innovation: a review and research agenda[J]. Technological Forecasting and Social Change, 188: 122264.\u003c/li\u003e\n\u003cli\u003eBai X, Han J, Ma Y, et al., 2022. ESG performance, institutional investors\u0026rsquo; preference and financing constraints: empirical evidence from China[J]. Borsa Istanbul Review, 22: S157-S168.\u003c/li\u003e\n\u003cli\u003eBalakrishnan K, Billings M B, Kelly B, et al., 2014. Shaping liquidity: on the causal effects of voluntary disclosure[J]. The Journal of Finance, 69(5): 2237-2278.\u003c/li\u003e\n\u003cli\u003eBarney J, 1991. Firm resources and sustained competitive advantage[J]. Journal of Management, 17(1): 99-120.\u003c/li\u003e\n\u003cli\u003eBegenau J, Farboodi M, Veldkamp L, 2018. Big data in finance and the growth of large firms[J]. Journal of Monetary Economics, 97: 71-87.\u003c/li\u003e\n\u003cli\u003eBelev B, 2003. Institutional investors in bulgarian corporate governance reform: obstacles or facilitators?[J]. Journal of World Business, 38(4): 361-374.\u003c/li\u003e\n\u003cli\u003eBelhadi A, Kamble ,Sachin, Fosso Wamba ,Samuel, et al., 2022. Building supply-chain resilience: an artificial intelligence-based technique and decision-making framework[J]. International Journal of Production Research, 60(14): 4487-4507.\u003c/li\u003e\n\u003cli\u003eBerente N, Gu B, Recker J, et al., 2021. Managing artificial intelligence[J]. MIS Quarterly.\u003c/li\u003e\n\u003cli\u003eBoot A, Hoffmann P, Laeven L, et al., 2021. Fintech: what\u0026rsquo;s old, what\u0026rsquo;s new?[J]. Journal of Financial Stability, 53: 100836.\u003c/li\u003e\n\u003cli\u003eBrickley J A, Lease R C, Smith C W, 1988. Ownership structure and voting on antitakeover amendments[J]. Journal of Financial Economics, 20: 267-291.\u003c/li\u003e\n\u003cli\u003eBushee B J, Core J E, Guay W, et al., 2010. The role of the business press as an information intermediary[J]. Journal of Accounting Research, 48(1): 1-19.\u003c/li\u003e\n\u003cli\u003eCao J, Dong D, Yue S, 2024. Institutional investors\u0026rsquo; site visits and firms\u0026rsquo; financial distress[J]. Research in International Business and Finance, 67: 102150.\u003c/li\u003e\n\u003cli\u003eChang Z, 2025. Influence of artificial intelligence on the financial labor market[J]. Finance Research Letters, 80: 107039.\u003c/li\u003e\n\u003cli\u003eChen K, Chen X, Wang Z ao, et al., 2024. Does artificial intelligence promote common prosperity within enterprises? \u0026mdash;evidence from chinese-listed companies in the service industry[J]. Technological Forecasting and Social Change, 200: 123180.\u003c/li\u003e\n\u003cli\u003eChen Y, Du L, Zhang B, et al., 2025. The impact of artificial intelligence on the sustainability of international trade enterprises[J]. International Review of Economics \u0026amp; Finance, 101: 104136.\u003c/li\u003e\n\u003cli\u003eCheng H, Huang D, Luo Y, 2020. Corporate disclosure quality and institutional investors\u0026rsquo; holdings during market downturns[J]. Journal of Corporate Finance, 60: 101523.\u003c/li\u003e\n\u003cli\u003eChishti M Z, Dogan E, Binsaeed R H, 2024. Can artificial intelligence and green finance affect economic cycles?[J]. Technological Forecasting and Social Change, 209: 123740.\u003c/li\u003e\n\u003cli\u003eChopra R, Sharma G D, Pereira V, 2024. Identifying bulls and bears? A bibliometric review of applying artificial intelligence innovations for stock market prediction[J]. Technovation, 135: 103067.\u003c/li\u003e\n\u003cli\u003eChoudhury T, 2025. US sectors and geopolitical risk: the investor\u0026rsquo;s perspective[J]. Finance Research Letters, 73: 106690.\u003c/li\u003e\n\u003cli\u003eChung H, Judge W Q, Li Y H, 2015. Voluntary disclosure, excess executive compensation, and firm value[J]. Journal of Corporate Finance.\u003c/li\u003e\n\u003cli\u003eChung K H, Zhang H, 2011. Corporate governance and institutional ownership[J]. Journal of Financial and Quantitative Analysis, 46(1): 247-273.\u003c/li\u003e\n\u003cli\u003eClarkson P M, Li Y, Richardson G D, et al., 2008. Revisiting the relation between environmental performance and environmental disclosure: an empirical analysis[J]. Accounting, Organizations and Society, 33(4-5): 303-327.\u003c/li\u003e\n\u003cli\u003eCohen L, Malloy C, Nguyen Q, 2020. Lazy prices[J]. The Journal of Finance, 75(3): 1371-1415.\u003c/li\u003e\n\u003cli\u003eCox P, Wicks P G, 2011. Institutional interest in corporate responsibility: portfolio evidence and ethical explanation[J]. Journal of Business Ethics.\u003c/li\u003e\n\u003cli\u003eCremers M, Pareek A, Sautner Z, 2020. Short-term investors, long-term investments, and firm value: evidence from russell 2000 index inclusions[J]. Management Science, 66(10): 4535-4551.\u003c/li\u003e\n\u003cli\u003eCulot G, Podrecca M, Nassimbeni G, 2024. Artificial intelligence in supply chain management: a systematic literature review of empirical studies and research directions[J]. Computers in Industry, 162: 104132.\u003c/li\u003e\n\u003cli\u003eCzarnitzki D, Fern\u0026aacute;ndez G P, Rammer C, 2023. Artificial intelligence and firm-level productivity[J]. Journal of Economic Behavior \u0026amp; Organization, 211: 188-205.\u003c/li\u003e\n\u003cli\u003eDas K Ch, Rani N, Bodhi R, et al., 2025. Do robots impact artificial intelligence (AI)-related employment? Evidence from a cross-national study[J]. International Journal of Manpower.\u003c/li\u003e\n\u003cli\u003eDing M, Gao Q, 2025. The impact of artificial intelligence technology application on total factor productivity in agricultural enterprises: evidence from China[J]. Economic Analysis and Policy, 86: 399-415.\u003c/li\u003e\n\u003cli\u003eDu K, Zhao Y, Mao R, et al., 2025. Natural language processing in finance: a survey[J]. Information Fusion, 115: 102755.\u003c/li\u003e\n\u003cli\u003eDuan D, Chen S, Feng Z, et al., 2023. Industrial robots and firm productivity[J]. Structural Change and Economic Dynamics, 67: 388-406.\u003c/li\u003e\n\u003cli\u003eDyck A, Lins K V, Roth L, et al., 2019. Do institutional investors drive corporate social responsibility? International evidence[J]. Journal of Financial Economics, 131(3): 693-714.\u003c/li\u003e\n\u003cli\u003eFan X, Wu Y, Zhou Y, et al., 2025. How does artificial intelligence shock affect labor income distribution? Evidence from China[J]. Pacific-Basin Finance Journal, 90: 102691.\u003c/li\u003e\n\u003cli\u003eFiorillo P, Meles A, Salerno D, et al., 2024. Geopolitical turmoil and investor green preference: evidence from the corporate bond market[J]. Journal of International Money and Finance, 149: 103218.\u003c/li\u003e\n\u003cli\u003eFlammer C, Toffel M W, Viswanathan K, 2021. Shareholder activism and firms\u0026rsquo; voluntary disclosure of climate change risks[J]. Strategic Management Journal, 42(10): 1850-1879.\u003c/li\u003e\n\u003cli\u003eFrancis B B, Hasan I, Zhu Y, 2021. The impact of political uncertainty on institutional ownership[J]. Journal of Financial Stability, 57: 100921.\u003c/li\u003e\n\u003cli\u003eFu Y, Shen Y, Song M, et al., 2024. Does artificial intelligence reduce corporate energy consumption? New evidence from China[J]. Economic Analysis and Policy, 83: 548-561.\u003c/li\u003e\n\u003cli\u003eGao L, Wang J, 2025. Can artificial intelligence reduce energy vulnerability? Evidence from an international perspective[J]. Energy Economics, 145: 108491.\u003c/li\u003e\n\u003cli\u003eGao Y, Liu S, Yang L, 2025. Artificial intelligence and innovation capability: a dynamic capabilities perspective[J]. International Review of Economics \u0026amp; Finance, 98: 103923.\u003c/li\u003e\n\u003cli\u003eGaver J J, Gaver K M, 1993. Additional evidence on the association between the investment opportunity set and corporate financing, dividend, and compensation policies[J]. Journal of Accounting and Economics, 16(1-3): 125-160.\u003c/li\u003e\n\u003cli\u003eGinesti G, Span\u0026ograve; R, Campa D, et al., 2025. Board composition, audit committee characteristics, and R\u0026amp;D expenditure[J]. European Management Journal, 43(1): 115-129.\u003c/li\u003e\n\u003cli\u003eGiordino D, Sestino A, Z\u0026agrave;mek D, et al., 2025. The impact of institutional ownership structure on corporate ESG performance: empirical evidence for the european capital market[J]. Review of Managerial Science.\u003c/li\u003e\n\u003cli\u003eGoh L, Liu X, Tsang A, 2020. Voluntary disclosure of corporate political spending[J]. Journal of Corporate Finance, 61: 101403.\u003c/li\u003e\n\u003cli\u003eGoldfarb A, Tucker C, 2019. Digital economics[J]. Journal of Economic Literature, 57(1): 3-43.\u003c/li\u003e\n\u003cli\u003eGoldsmith-Pinkham P, Sorkin I, Swift H, 2020. Bartik instruments: what, when. Why, and how[J]. American Economic Review, 110(8): 2586-2624.\u003c/li\u003e\n\u003cli\u003eDeb G S, 2018. Institutional investors and firm characteristics: new evidence from India[J]. Research in International Business and Finance, 46: 30-42.\u003c/li\u003e\n\u003cli\u003eHaefner N, Wincent J, Parida V, et al., 2021. Artificial intelligence and innovation management: a review, framework, and research agenda✰[J]. Technological Forecasting and Social Change, 162: 120392.\u003c/li\u003e\n\u003cli\u003eHan Q, Li C, Jin Y, 2025. The impact of intellectual property protection on the development of artificial intelligence in enterprises[J]. International Review of Financial Analysis: 104179.\u003c/li\u003e\n\u003cli\u003eHao (Grace) Qing, Li A, 2024. Economic policy uncertainty and institutional portfolio investment[J]. Financial Management, n/a(n/a).\u003c/li\u003e\n\u003cli\u003eHaushalter D, Klasa S, Maxwell W, 2007. The influence of product market dynamics on a firm\u0026rsquo;s cash holdings and hedging behavior☆[J]. Journal of Financial Economics, 84(3): 797-825.\u003c/li\u003e\n\u003cli\u003eHe G, Li Z, Yu L, et al., 2025. Does commercial reform embracing digital technologies mitigate stock price crash risk?[J]. Journal of Corporate Finance, 91: 102741.\u003c/li\u003e\n\u003cli\u003eHeckman J J, 1979. Sample selection bias as a specification error[J]. Econometrica, 47(1): 153-161.\u003c/li\u003e\n\u003cli\u003eHong H, Kubik J D, Stein J C, 2008. The only game in town: stock-price consequences of local bias$[J]. Journal of Financial Economics.\u003c/li\u003e\n\u003cli\u003eHuang J, Han F, Li Y, 2025. The impact of institutional investors\u0026rsquo; ESG concerns on corporate ESG disclosure: evidence from site visits[J]. Finance Research Letters, 76: 106957.\u003c/li\u003e\n\u003cli\u003eHuang M, Li Q, Li B, 2025. From algorithms to green growth: can artificial intelligence drive enterprise energy transformation?[J]. Economic Analysis and Policy, 85: 1846-1866.\u003c/li\u003e\n\u003cli\u003eHuang Q, Li Y, Lin M, et al., 2022. Natural disasters, risk salience, and corporate ESG disclosure[J]. Journal of Corporate Finance, 72: 102152.\u003c/li\u003e\n\u003cli\u003eIacucci M, Santacroce G, Yasuharu M, et al., 2025. Artificial intelligence\u0026ndash;driven personalized medicine: transforming clinical practice in inflammatory bowel disease[J]. Gastroenterology: S0016508525004949.\u003c/li\u003e\n\u003cli\u003eJia H, Che W, 2025. Institutional investor stability, executive equity incentives, and corporate innovation[J]. Finance Research Letters, 83: 107691.\u003c/li\u003e\n\u003cli\u003eJiang C X, Kim J C, Zhou D, 2011. Liquidity, analysts, and institutional ownership[J]. International Review of Financial Analysis, 20(5): 335-344.\u003c/li\u003e\n\u003cli\u003eJiang X, Yuan Q, 2018. Institutional investors\u0026rsquo; corporate site visits and corporate innovation[J]. Journal of Corporate Finance, 48: 148-168.\u003c/li\u003e\n\u003cli\u003eJin C, Monfort A, Chen F, et al., 2024. Institutional investor ESG activism and corporate green innovation against climate change: exploring differences between digital and non-digital firms[J]. Technological Forecasting and Social Change, 200: 123129.\u003c/li\u003e\n\u003cli\u003eJohnson P C, Laurell C, Ots M, et al., 2022. Digital innovation and the effects of artificial intelligence on firms\u0026rsquo; research and development \u0026ndash; automation or augmentation, exploration or exploitation?[J]. Technological Forecasting and Social Change, 179: 121636.\u003c/li\u003e\n\u003cli\u003eJoseph J, Wilson A J, 2018. The growth of the firm: an attention-based view[J]. Strategic Management Journal, 39(6): 1779-1800.\u003c/li\u003e\n\u003cli\u003eKacperczyk M T, Nosal J, Wang T, 2022. Global volatility and firm-level capital flows[J]. SSRN Electronic Journal.\u003c/li\u003e\n\u003cli\u003eKoenker R, Bassett G, 1978. Regression quantiles[J]. Econometrica, 46(1): 33-50.\u003c/li\u003e\n\u003cli\u003eKong D, Lin C, Liu S, et al., 2021. Whose money is smart? Individual and institutional investors\u0026rsquo; trades based on analyst recommendations[J]. Journal of Empirical Finance, 62: 234-251.\u003c/li\u003e\n\u003cli\u003eKrueger P, Sautner Z, Starks L T, 2019. The importance of climate risks for institutional investors[J].\u003c/li\u003e\n\u003cli\u003eLai S, Li X, Liu S, et al., 2022. Institutional investors\u0026rsquo; site visits and corporate employment decision-making[J]. Journal of Contemporary Accounting \u0026amp; Economics, 18(3): 100332.\u003c/li\u003e\n\u003cli\u003eLaksmana I, Harjoto M A, Kim H, 2023. Managing disclosure of political risk: the case of socially responsible firms[J]. Journal of Business Research, 154: 113366.\u003c/li\u003e\n\u003cli\u003eLee C C, Qin S, Li Y, 2022. Does industrial robot application promote green technology innovation in the manufacturing industry?[J]. Technological Forecasting and Social Change, 183: 121893.\u003c/li\u003e\n\u003cli\u003eLewellen J, Lewellen K, 2022. Institutional investors and corporate governance: the incentive to Be engaged[J]. The Journal of Finance, 77(1): 213-264.\u003c/li\u003e\n\u003cli\u003eLi C, Xu Y, Zheng H, et al., 2023. Artificial intelligence, resource reallocation, and corporate innovation efficiency: evidence from China\u0026rsquo;s listed companies[J]. Resources Policy, 81: 103324.\u003c/li\u003e\n\u003cli\u003eLi G, Ren K, Qiao Y, et al., 2024. From framing to priming: how does media coverage promote ESG preferences of institutional investors[J]. Emerging Markets Review, 63: 101220.\u003c/li\u003e\n\u003cli\u003eLi G, Shao Y, 2023. How do top management team characteristics affect digital orientation? Exploring the internal driving forces of firm digitalization[J]. Technology in Society, 74: 102293.\u003c/li\u003e\n\u003cli\u003eLi H, Lu Z, Zhang Z, et al., 2025. How does artificial intelligence affect manufacturing firms\u0026rsquo; energy intensity?[J]. Energy Economics, 141: 108109.\u003c/li\u003e\n\u003cli\u003eLi H, Wu H, Rao J, 2025. Impact of artificial intelligence on corporate green transformation[J]. Finance Research Letters, 80: 107427.\u003c/li\u003e\n\u003cli\u003eLi J, Li M, Wang X, et al., 2021. Strategic directions for AI: the role of cios and boards of directors[J]. MIS Quarterly, 45(3): 1603-1643.\u003c/li\u003e\n\u003cli\u003eLi J, Ma S, Qu Y, et al., 2023. The impact of artificial intelligence on firms\u0026rsquo; energy and resource efficiency: empirical evidence from China[J]. Resources Policy, 82: 103507.\u003c/li\u003e\n\u003cli\u003eLi J, Wu T, Hu B, et al., 2025. Artificial intelligence and corporate ESG performance[J]. International Review of Financial Analysis, 102: 104036.\u003c/li\u003e\n\u003cli\u003eLi W, Lu X, 2016. Institutional interest, ownership type, and environmental capital expenditures: evidence from the most polluting chinese listed firms[J]. Journal of Business Ethics, 138(3): 459-476.\u003c/li\u003e\n\u003cli\u003eLi Y, Goodell J W, Shen D, 2023. Market reaction to climate risk report disclosures: the roles of investor attention and sentiment[J]. Finance Research Letters, 58: 104290.\u003c/li\u003e\n\u003cli\u003eLi Y, Zhang T, Tao Y, et al., 2025. Technology-based boards and bank operational efficiency: Mediating effects of risk-taking[J]. Finance Research Letters, 76: 106924.\u003c/li\u003e\n\u003cli\u003eLin F, Wu C M, Fang T Y, et al., 2014. The relations among accounting conservatism, institutional investors and earnings manipulation[J]. Economic Modelling, 37: 164-174.\u003c/li\u003e\n\u003cli\u003eLin Y C, Padliansyah R, Lu Y H, et al., 2025. Bankruptcy prediction: integration of convolutional neural networks and explainable artificial intelligence techniques[J]. International Journal of Accounting Information Systems, 56: 100744.\u003c/li\u003e\n\u003cli\u003eLiu J, Xiong X, Gao Y, et al., 2023. The impact of institutional investors on ESG: evidence from China[J]. Accounting \u0026amp; Finance, 63(S2): 2801-2826.\u003c/li\u003e\n\u003cli\u003eLiu R, Luo W, Su R, 2025. How does cultural diversity influence corporate AI development?[J]. Finance Research Letters, 81: 107506.\u003c/li\u003e\n\u003cli\u003eLiu Y, Jin X M, Ly K C, et al., 2024. The relationship between heterogeneous institutional investors\u0026rsquo; shareholdings and corporate ESG performance: evidence from China[J]. Research in International Business and Finance, 71: 102457.\u003c/li\u003e\n\u003cli\u003eLiu Y, Shen F, Guo J, et al., 2025. Can artificial intelligence technology improve companies\u0026rsquo; capacity for green innovation? Evidence from listed companies in China[J]. Energy Economics, 143: 108280.\u003c/li\u003e\n\u003cli\u003eLiu Y, Wang B, Song Z, 2025. Promoting or inhibiting: the impact of artificial intelligence application on corporate environmental performance[J]. International Review of Financial Analysis, 97: 103872.\u003c/li\u003e\n\u003cli\u003eLuqman A, Zhang Q, Talwar S, et al., 2024. Artificial intelligence and corporate carbon neutrality: a qualitative exploration[J]. Business Strategy and the Environment, 33(5): 3986-4003.\u003c/li\u003e\n\u003cli\u003eLyssimachou D, Bilinski P, 2023. Does corporate social responsibility affect the institutional ownership of firms in the hospitality and tourism industry?[J]. Tourism Economics, 29(4): 853-879.\u003c/li\u003e\n\u003cli\u003eMarshall A, Rao S, Roy P P, et al., 2022. Mandatory corporate social responsibility and foreign institutional investor preferences[J]. Journal of Corporate Finance, 76: 102261.\u003c/li\u003e\n\u003cli\u003eMcCahery J A, Sautner Z, Starks L T, 2016. Behind the scenes: the corporate governance preferences of institutional investors[J]. Journal of Finance, 71(6): 2905-2932.\u003c/li\u003e\n\u003cli\u003eNath P C, Mishra A K, Sharma R, et al., 2024. Recent advances in artificial intelligence towards the sustainable future of agri-food industry[J]. Food Chemistry, 447: 138945.\u003c/li\u003e\n\u003cli\u003eNguyen Q P, Vo D H, 2022. Artificial intelligence and unemployment:an international evidence[J]. Structural Change and Economic Dynamics, 63: 40-55.\u003c/li\u003e\n\u003cli\u003eOcasio W, 1997. Towards an attention-based view of the firm[J]. Strategic Management Journal, 18(S1): 187-206.\u003c/li\u003e\n\u003cli\u003ePanon L, 2022. Labor share, foreign demand and superstar exporters[J]. Journal of International Economics, 139: 103678.\u003c/li\u003e\n\u003cli\u003ePersad K, Xu B, Greening P, 2024. Institutional investor behaviour and the energy transition: a complexity framework for accelerating sustainable finance from UK investors[J]. Energy Economics: 107444.\u003c/li\u003e\n\u003cli\u003ePinski M, Hofmann T, Benlian A, 2024. AI literacy for the top management: an upper echelons perspective on corporate AI orientation and implementation ability[J]. Electronic Markets, 34(1): 24.\u003c/li\u003e\n\u003cli\u003ePolireddi N S A, 2024. An effective role of artificial intelligence and machine learning in banking sector[J]. Measurement: Sensors, 33: 101135.\u003c/li\u003e\n\u003cli\u003eQu G, Jing H, 2025. Is new technology always good? Artificial intelligence and corporate tax avoidance: evidence from China[J]. International Review of Economics \u0026amp; Finance, 98: 103949.\u003c/li\u003e\n\u003cli\u003eRaimondo C, 2019. The media and the financial markets: a review[J]. Asia-pacific Journal of Financial Studies, 48(2): 155-184.\u003c/li\u003e\n\u003cli\u003eRammer C, Fern\u0026aacute;ndez G P, Czarnitzki D, 2022. Artificial intelligence and industrial innovation: evidence from german firm-level data[J]. Research Policy, 51(7): 104555.\u003c/li\u003e\n\u003cli\u003eRees L, Sharp N, Twedt B, 2015. Who\u0026rsquo;s heard on the street? Determinants and consequences of financial analyst coverage in the business press[J]. Review of Accounting Studies, 20(1): 173-209.\u003c/li\u003e\n\u003cli\u003eRen B, Qiu Z, Liu B, 2025. Supply chain decarbonisation effects of artificial intelligence: evidence from China[J]. International Review of Economics \u0026amp; Finance, 101: 104198.\u003c/li\u003e\n\u003cli\u003eRoberts D L, Candi M, 2024. Artificial intelligence and innovation management: charting the evolving landscape[J]. Technovation, 136: 103081.\u003c/li\u003e\n\u003cli\u003eRoberts H, Cowls J, Morley J, et al., 2021. The chinese approach to artificial intelligence: an analysis of policy, ethics, and regulation[J]. AI And Society, 36(1): 59-77.\u003c/li\u003e\n\u003cli\u003eRoller S, 2021. Institutional investors\u0026rsquo; sustainability policies and the outcomes for the healthcare sector portfolios[J]. Health Policy, 125(10): 1367-1376.\u003c/li\u003e\n\u003cli\u003eSaba C S, Ngepah N, 2024. The impact of artificial intelligence (AI) on employment and economic growth in BRICS: does the moderating role of governance matter?[J]. Research in Globalization, 8: 100213.\u003c/li\u003e\n\u003cli\u003eSaci F, Jasimuddin S M, 2021. Does the research done by the institutional investors affect the cost of equity capital?[J]. Finance Research Letters, 41: 101834.\u003c/li\u003e\n\u003cli\u003eSkare M, Gavurova B, Blažević Burić S, 2024. Artificial intelligence and wealth inequality: a comprehensive empirical exploration of socioeconomic implications[J]. Technology in Society, 79: 102719.\u003c/li\u003e\n\u003cli\u003eSlimani S, Omri A, Ben Jabeur S, 2025. When and how does artificial intelligence impact environmental performance?[J]. Energy Economics, 148: 108643.\u003c/li\u003e\n\u003cli\u003eSong Y, 2024. Institutional investors\u0026rsquo; corporate site visits and firm-level climate change risk disclosure[J]. International Review of Financial Analysis.\u003c/li\u003e\n\u003cli\u003eSong Y, Xian R, 2024. Institutional investors\u0026rsquo; corporate site visits and firm-level climate change risk disclosure[J]. International Review of Financial Analysis, 93: 103145.\u003c/li\u003e\n\u003cli\u003eStiglitz J E, 2000. The contributions of the economics of information to twentieth century economics[J]. The Quarterly Journal of Economics, 115(4): 1441-1478.\u003c/li\u003e\n\u003cli\u003eSullivan Y, Fosso Wamba S, 2024. Artificial intelligence and adaptive response to market changes: a strategy to enhance firm performance and innovation[J]. Journal of Business Research, 174: 114500.\u003c/li\u003e\n\u003cli\u003eSun Z, Che S, Wang J, 2024. Deconstruct artificial intelligence\u0026rsquo;s productivity impact: a new technological insight[J]. Technology in Society, 79: 102752.\u003c/li\u003e\n\u003cli\u003eTao W, Weng S, Chen X, et al., 2024. Artificial intelligence-driven transformations in low-carbon energy structure: evidence from China[J]. Energy Economics, 136: 107719.\u003c/li\u003e\n\u003cli\u003eTeece D J, 2007. Explicating dynamic capabilities: the nature and microfoundations of (sustainable) enterprise performance[J]. Strategic Management Journal, 28(13): 1319-1350.\u003c/li\u003e\n\u003cli\u003eThompson L, 2025. Human capital disclosures and institutional ownership[J]. Finance Research Letters, 71: 106384.\u003c/li\u003e\n\u003cli\u003eTian H, Zhao L, Yunfang L, et al., 2023. Can enterprise green technology innovation performance achieve \u0026ldquo;corner overtaking\u0026rdquo; by using artificial intelligence?\u0026mdash;evidence from chinese manufacturing enterprises[J]. Technological Forecasting and Social Change, 194: 122732.\u003c/li\u003e\n\u003cli\u003eTie, Y., Liu, Y., 2025. External demand shocks and changes in labor income shares. Finance \u0026amp; Trade Economics,46(7):56-72.\u003c/li\u003e\n\u003cli\u003eVeganzones D, S\u0026eacute;verin E, 2025. Earnings management visualization and prediction using machine learning methods[J]. International Journal of Accounting Information Systems, 56: 100743.\u003c/li\u003e\n\u003cli\u003eWang C, Nazar R, Ali S, et al., 2025. Eco-friendly algorithms: artificial intelligence and green finance in european intelligent nations[J]. Technology in Society: 102960.\u003c/li\u003e\n\u003cli\u003eWang J, Wang A, Luo K, et al., 2024. Can artificial intelligence improve enterprise environmental performance: evidence from China[J]. Journal of Environmental Management, 370: 123079.\u003c/li\u003e\n\u003cli\u003eWang L, Chen Y, 2025. Artificial intelligence and corporate investment efficiency: evidence from China[J]. Emerging Markets Review, 68: 101314.\u003c/li\u003e\n\u003cli\u003eWang Q, Zhang F, Li R, et al., 2024. Does artificial intelligence promote energy transition and curb carbon emissions? The role of trade openness[J]. Journal of Cleaner Production, 447: 141298.\u003c/li\u003e\n\u003cli\u003eWang S, Sun Z, Wang H, et al., 2025. Enhancing student acceptance of artificial intelligence-driven hybrid learning in business education: interaction between self-efficacy, playfulness, emotional engagement, and university support[J]. The International Journal of Management Education, 23(2): 101184.\u003c/li\u003e\n\u003cli\u003eWang S, Zhang Y, Xiao Y, et al., 2025. Artificial intelligence policy frameworks in China, the european union and the United States: an analysis based on structure topic model[J]. Technological Forecasting and Social Change, 212: 123971.\u003c/li\u003e\n\u003cli\u003eWang X, Wei S, Zhu X, 2024. Economic policy uncertainty and heterogeneous institutional investor horizons[J]. Review of Quantitative Finance and Accounting, 62(1): 39-67.\u003c/li\u003e\n\u003cli\u003eWang Y, Zhang R, Yao K, et al., 2024. Does artificial intelligence affect the ecological footprint? \u0026ndash;evidence from 30 provinces in China[J]. Journal of Environmental Management, 370: 122458.\u003c/li\u003e\n\u003cli\u003eWang, Z., Yang, G., Xiong, H., 2023. Corporate Innovation Focus and Institutional Investor Ownership: Evidence from Listed Firm Patent Classification Number Data. South China Journal of Economics (7): 42-62.\u003c/li\u003e\n\u003cli\u003eWen H, Shang J, Nghiem X H, 2025. Does artificial intelligence matter for the population aging-inclusive growth nexus? International evidence[J]. Telecommunications Policy, 49(4): 102932.\u003c/li\u003e\n\u003cli\u003eWeng S, Rong Z, Yu L, 2024. Institutional site visits and corporate innovation: the role of analyst coverage[J]. Pacific-basin Finance Journal, 85: 102354.\u003c/li\u003e\n\u003cli\u003eWernerfelt B, 1984. A resource-based view of the firm[J]. Strategic Management Journal, 5(2): 171-180.\u003c/li\u003e\n\u003cli\u003eWu B, Chen H, Shi Y, 2025. Influence of artificial intelligence development on supply chain diversification[J]. Finance Research Letters, 78: 107210.\u003c/li\u003e\n\u003cli\u003eWu Q, Xie T, Lu J, 2025. Does institutional investors\u0026rsquo; environmental attention influence corporate environmental investment?[J]. Economic Analysis and Policy, 85: 1625-1644.\u003c/li\u003e\n\u003cli\u003eWu Y, Lin Z, Zhang Q, et al., 2024. Artificial intelligence, wage dynamics, and inequality: empirical evidence from chinese listed firms[J]. International Review of Economics \u0026amp; Finance, 96: 103739.\u003c/li\u003e\n\u003cli\u003eWu Y, Zhang J, Cai X, 2025. Impact of regional artificial intelligence development on corporate environmental information[J]. Finance Research Letters, 80: 107413.\u003c/li\u003e\n\u003cli\u003eXia Y, Xie L, Chen X, et al., 2025. Artificial intelligence and corporate fraud: evidence from China[J]. Economic Analysis and Policy, 86: 1391-1408.\u003c/li\u003e\n\u003cli\u003eXiao H, 2023. Institutional investors\u0026rsquo; corporate site visits and corporate investment efficiency[J]. International Review of Finance, 23(2): 359-392.\u003c/li\u003e\n\u003cli\u003eXu X, Wan Q, Ma J, et al., 2025. What drives institutional investors\u0026rsquo; site visit decisions: the role of local business environment in China[J]. International Review of Financial Analysis, 103: 104204.\u003c/li\u003e\n\u003cli\u003eYahia N B, Chalwati A, Hmaied D, et al., 2023. Do foreign institutions avoid investing in poorly CSR-performing firms?[J]. Journal of Banking \u0026amp; Finance, 157: 107029.\u003c/li\u003e\n\u003cli\u003eYan X (Sterling), Zhang Z, 2009. Institutional investors and equity returns: are short-term institutions better informed?[J]. Review of Financial Studies, 22(2): 893-924.\u003c/li\u003e\n\u003cli\u003eYang Y, An R, Song J, 2025. Impact of enterprise artificial intelligence on social responsibility: evidence from text analysis[J]. Finance Research Letters, 75: 106868.\u003c/li\u003e\n\u003cli\u003eYang Y, Yi C, Li H, et al., 2025. An analysis on the role of artificial intelligence in green supply chains[J]. Technological Forecasting and Social Change, 217: 124169.\u003c/li\u003e\n\u003cli\u003eYao, J., Zhang, K., Guo, L., et al., 2024. How does artificial intelligence improve firm productivity? Based on the perspective of labor skill structure adjustment. Management World 40 (2), 101\u0026ndash;116.\u003c/li\u003e\n\u003cli\u003eYuan R, Wen W, 2018. Managerial foreign experience and corporate innovation[J]. Journal of Corporate Finance, 48: 752-770.\u003c/li\u003e\n\u003cli\u003eZhang C, Yang J, 2024. Artificial intelligence and corporate ESG performance[J]. International Review of Economics \u0026amp; Finance, 96: 103713.\u003c/li\u003e\n\u003cli\u003eZhang J, 2025. Artificial intelligence and green innovation: investigating the effects of executive pay and firm age[J]. Research in International Business and Finance, 77: 102953.\u003c/li\u003e\n\u003cli\u003eZhang J, Ye Y, 2023. Does the research meeting affect the shareholding ratio of institutional investors in listed companies? Empirical evidence from China[J]. China Finance Review International, 13(4): 714-733.\u003c/li\u003e\n\u003cli\u003eZhang P, Ma B, Chi C, 2025. Institutional investors and ESG performance: Evidence from China[J]. Economic Analysis and Policy, 86: 1159-1181.\u003c/li\u003e\n\u003cli\u003eZhao C, Li Y, Liu Z, et al., 2024. Artificial intelligence and carbon emissions inequality: Evidence from industrial robot application[J]. Journal of Cleaner Production, 438: 140817.\u003c/li\u003e\n\u003cli\u003eZhao L, Li N, Wu Y, 2023. Institutional investors\u0026rsquo; site visits, information asymmetry, and investment efficiency[J]. International Review of Financial Analysis, 88: 102674.\u003c/li\u003e\n\u003cli\u003eZhong K, Song L, 2025. Artificial intelligence adoption and corporate green innovation capability[J]. Finance Research Letters, 72: 106480.\u003c/li\u003e\n\u003cli\u003eZhou X, Li G, Wang Q, et al., 2025. Artificial intelligence, corporate information governance and ESG performance: quasi-experimental evidence from China[J]. International Review of Financial Analysis, 102: 104087.\u003c/li\u003e\n\u003cli\u003eZou J, Zhong X, Gong C, et al., 2025. Examining institutional investor preferences: the influence of ESG ratings on stock holding in China\u0026rsquo;s stock market[J]. Research in International Business and Finance, 73: 102609.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Footnotes","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003e International Data Corporate (IDC), founded in 1964, is a global professional consulting services organization affiliated with International Data Group (IDG).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e International Business Machines Corporation (IBM), headquartered in New York, and founded in the United States in 1911, is the world's largest information technology and business solutions company.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e The 2017 China Artificial Intelligence Industry Special Research Report states that the surge in the development of artificial intelligence mainly began with the introduction of deep learning algorithms in 2006.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"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":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"artificial intelligence information disclosure, institutional investor shareholding, on-site visits, information transparency, innovation","lastPublishedDoi":"10.21203/rs.3.rs-7134593/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7134593/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eArtificial intelligence (AI) is a significant driver propelling the new wave of technological revolution and industrial transformation, exerting a profound influence across various sectors. This paper examines the relationship between AI information disclosure and changes in institutional investor shareholdings, utilizing data from Chinese A-share companies listed on the Shanghai and Shenzhen stock exchanges between 2007 and 2023. The results indicate that enhanced AI information disclosure is associated with increased institutional investor ownership. This conclusion remains valid following robustness tests that include substituting dependent and independent variables, altering clustering levels, excluding municipal samples, and employing quantile regression. Heterogeneity analyses further reveal that the effect is notably stronger in samples post-2011, within manufacturing firms, in environments with higher competition, and among firms with directors possessing a strong technological background. The mechanism analysis reveals that increased AI information disclosure prompts institutional investors to conduct on-site visits, enhancing information transparency and bolstering innovation capabilities. Further, heightened AI information disclosure is particularly significant in encouraging shareholdings by long-term and resilient institutional investors. This paper presents the first micro-level evidence on the impact of AI disclosure on institutional investor shareholdings, offering theoretical and practical implications for refining investment strategies of institutional investors in the capital market.\u003c/p\u003e\n\u003cp\u003eJEL Classification G10 \u0026nbsp;G32\u003c/p\u003e","manuscriptTitle":"AI Disclosure as an Institutional Investment Catalyst: Evidence from China","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-21 09:31:16","doi":"10.21203/rs.3.rs-7134593/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"e5096adb-f986-4e69-acdb-9ea543b78f87","owner":[],"postedDate":"August 21st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-07T06:38:23+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-21 09:31:16","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7134593","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7134593","identity":"rs-7134593","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}