Using AI-driven services to reduce food waste in restaurants: The role of social presence and technology agency level

Using AI-driven services to reduce food waste in restaurants: The role of social presence and technology agency level | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Using AI-driven services to reduce food waste in restaurants: The role of social presence and technology agency level Jing Zhang, Ivan Ka Wai Lai, Xiaohong Wu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8843012/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 This study investigates how AI-driven services can enhance consumers’ moral norms to reduce food waste in Chinese hot pot and barbecue buffets. A between-subjects 2 social presence in ordering system (high vs. low) x 2 technology agency level (high vs. low) factorial experimental design was used. Two experiments were performed via face-to-face questionnaire surveys with 181 and 363 respondents, respectively. The results indicate that high social presence (via AI-assisted ordering) enhances consumers’ moral norms and their intention to reduce food waste (IRFW). The high agency level (providing dish suggestions) moderates the effect of ordering system on moral norms and IRFW, and the perception of moral norms mediates the interactive effects of social presence and technology agency level on IRFW. This study contributes to the field of hospitality technology research by examining the role of AI-driven services in reducing food waste in restaurants. Business and commerce/Business and management Social science/Business and management Earth and environmental sciences/Environmental social sciences Biological sciences/Psychology Social science/Psychology AI-driven services social presence technology agency food waste reduction moral norms Figures Figure 1 Figure 2 Figure 3 1. Introduction Food waste is a substantial environmental concern worldwide. The United Nations Sustainable Development Goals (UN SDGs) call for cutting global per capita food waste at the retail and consumer levels in half by 2030. Reducing food waste is a challenging task, a global focus, and has attracted great recent research interest (Filimonau et al., 2023 ). Consequently, many studies on reducing food waste have been conducted, but most have been carried out in Western countries (Dhir et al., 2020 ). In China, the catering industry is facing a significant food waste problem, with consumers throwing away between 17 and 18 million tons of food annually, an amount that could feed 30–50 million people (Khan, 2018 ). Despite the introduction of the “Clean Your Plate Campaign” (CYPC) in China over a decade ago, which heavily relies on personal ethics (Wang et al., 2022 ), progress remains limited, particularly in unique dining formats such as hot pot and barbecue buffets. These popular, self-cooking experiences, fundamentally different from Western pre-prepared buffets (Zhao et al., 2020 ), exacerbate waste through behaviours like over-selecting uncooked ingredients or leaving cooked food uneaten (Pang et al, 2023 ). As more and more Chinese hot pot and barbecue restaurants adopt self-ordering systems, customers can order uncooked ingredients by simply clicking on the menu. This quick operation makes it easier for customers to overorder. Given the popularity of self-ordering systems and the global spread of these restaurants, there is a need to investigate how to reduce consumers’ food waste in hot pot and barbecue restaurants by using technologies, especially in China. Artificial intelligence (AI) has been widely used in enterprise green management (Che et al., 2026 ). Leading businesses in the hospitality industry have begun leveraging AI to reduce food waste (Clark et al., 2025 ). For example, Marriott Hotel has successfully reduced food waste through the integration of the Winnow AI platform, which provides real-time insights into food consumption and waste, empowering the hotel to make smarter decisions around ordering, menu planning, and portion size (Adams, 2024 ). Bi et al.’s ( 2025 ) study on the impact of artificial intelligence on consumers' willingness to purchase healthy food and its role in reducing food waste. While current AI applications are primarily focused on back-end operations, AI can excel in front-end operations and even perform better in certain tasks. For instance, AI waiters in a restaurant can track customers’ orders and provide personalised portion guidance like human staff. However, recent hospitality research on AI to reduce food waste focuses on its back-end functions (e.g. Adams, 2024 ; Clark et al., 2025 ). There is a lack of empirical research on how AI-driven service models can promote sustainable dining behaviours, especially in high-waste environments like Chinese hot pot and barbecue buffets. Different technologies may generate various senses of social presence, where technology is perceived as a sentient, social actor rather than a tool (Kim et al., 2022 ). Recently, most restaurants adopted self-ordering systems, and Chinese hot pot and barbecue buffet restaurants are no exception. Self-ordering systems, typically scan QR codes, are functionally efficient but inherently low in social presence. When interacting with such a low social presence interface, customers do not have the sense of the presence of others; therefore, customers may choose a lot of ingredients without thinking carefully about whether to eat them all. However, AI-assisted ordering systems can communicate with customers just like a human waiter, which automatically evokes a sense of social presence (Bai et al., 2024 ). Norms activation theory posits that individual behaviour is influenced by the activation of their personal norms (Schwartz, 1977 ). Perceptions of AI as a human waiter can activate a sense of moral norms, preventing customers from ordering too much food and reducing waste. In addition, technology has different levels of agency, defined as its capability to act independently on behalf of the customer (Adams et al., 2022 ). Systems with a low level of agency merely execute customer instructions, lacking effective guidance. In a restaurant setting, providing dish suggestions based on the number of diners and their preferences showcases a high level of agency. It shifts the technology from a passive tool to an active participant in diners’ decision-making (Anderson et al., 2024 ), as they become active participants in shaping shared spatial experiences with humans, even shaping social norms (Pischetola et al., 2021 ). Therefore, the technology agency level in providing disk suggestions may leverage and amplify the effect of social presence. Therefore, customers will have a higher sense of moral norms and will have a higher intention not to order too many dishes to reduce food waste when the system providing disk suggestions, whether using a self-ordering system or an AI-assisted ordering system. This study aims to investigate the ways of how AI-driven services can enhance consumers’ moral norms to reduce food waste in Chinese hot pot and barbecue buffets. This study particularly answers below sub-questions: (1) Do the perceptions of social presence in the ordering system (high vs. low) influence consumers’ moral norms and increase their IRFW? (2) Do moral norms mediate the relationship between ordering system and IRFW? and (3) Do the perceptions of technology agency level (high vs. low) moderate the effect of ordering system on customers’ moral norms and IRFW? To answer the above sub-questions, this study employs a factorial experimental design and consists of two experiments. Study 1 is a one-factor between-groups experimental design, and Study 2 is a 2 x 2 factorial between-groups experimental design. Firstly, this study contributes to environmental protection research by adopting AI technologies to drive customers’ environmental protection behaviours in the front-end food service context. Based on social presence theory, agency theory, and moral theory, this study explored how two types of features (constructing social presence and increasing technology agency level) of AI technologies implicitly induce moral norms, showing researchers and the catering business a new and friendly way to influence consumer sustainable behaviour. It also enriches the literature on AI-assisted services by studying how these two types of AI characteristics interact to influence customer responses. Finally, previous studies on reducing restaurant food waste mainly proposed legal approaches (Szulecka et al., 2024 ) and campaign approaches (Wang et al., 2022 ) and using explicit informational interventions as stimuli under human services to test the effect of social norms (e.g. Budovska et al., 2020 ). This study provides a novel perspective on inducing people’s moral norms to address food waste in China, especially in the Chinese hot pot and barbecue buffet context. In addition, practical suggestions are provided to catering businesses on using AI-driven services to reduce food waste. 2. Literature review 2.1 Food waste and AI in food waste reduction Food waste refers to the edible portion of food that is discarded or not consumed. Food waste is an increasing issue (Lei, Agyeiwaah et al., 2024 ), leading to notable environmental, economic, and social consequences (Stöckli et al., 2018 ). Many countries have introduced anti-food waste laws to address food waste issues. For example, China legislated its first anti-food waste law (the Anti-food Waste Law of the People's Republic of China) in 2021 (SWITCH-Asia, 2024 ). In addition, numerous campaigns and programs have emerged to reduce food waste, such as the “Love Food, Hate Waste” campaign in the UK and the CYPC in China. However, the effectiveness of most campaigns is short, including China’s CYPC. According to the report on food waste by country in 2025, the food waste per capita in 2024 in China was about 76kg, which was about a 19% increase from 2021 (64kg) (World Population Review, 2025 ). Researchers have explored a variety of strategies to reduce food waste, including using AI. AI is a technology that allows electronic devices to imitate human actions (Bowen & Morosan, 2018 ), which has the potential to revolutionise the restaurant industry by enhancing sustainable practices (Narayanan et al., 2024 ) and reduce waste (Che et al., 2026 ). AI has been widely utilised in the hospitality and tourism industry and has attracted significant research interests, including employing AI to reduce food waste. For example, Onyeaka et al. ( 2025 ) conducted a systematic review of AI technologies (e.g. machine learning, predictive analytics) for food spoilage prediction across supply chain stages, highlighting their potential to reduce waste while addressing data and ethical challenges in implementation. Narayanan et al. ( 2024 ) proposed a holographic “AI+virtual reality (VR)” ordering system to enhance restaurant operational efficiency, reduce paper waste, and improve customer experience through digital menus, wireless orders, and intelligent recommendation algorithms, thereby promoting environmental sustainability in the catering industry. Despite research interest in AI in the hospitality sector, much of the existing literature is either conceptual (e.g. Narayanan et al, 2024 ) or literature review (e.g. Onyeaka et al., 2025 ), with empirical studies primarily centred on customers’ adoption (e.g. Alam et al., 2025 ). Consequently, there is a lack of research examining whether the utilisation of AI-driven services can lead customers to have an intention to reduce wasting food in restaurants compared to self-ordering services, and the underlying mechanisms. 2.2 Social presence and technology agency Social presence is the sensation of being with others (Biocca et al., 2003 ). Since people respond socially to both human and computer-controlled entities (Nowak & Biocca, 2023). With the technological advancements, the concept of social presence has evolved from human interaction to human-technology interaction (van Doorn et al., 2017 ). Advanced technology, such as a robot, enables humans to experience automated social presence by creating a feeling of being in the presence of another social entity (van Doorn et al., 2017 ). This feeling of social presence during technology interaction can be triggered by technological factors (Kim et al., 2022 ), such as humanisation (Casaló et al., 2025 ). For example, the advancements in AI technology have the potential to evoke a strong sense of social presence during the interaction between customers and AI (Kim et al., 2022 ). Social presence enhances positive mediated virtual experience (Biocca et al., 2003 ), playing a critical role in understanding how humans interact with machine agents (Kim et al., 2022 ). However, whether the role of social presence in the restaurant ordering systems can shape customers’ intention to reduce food waste remains unclear. Technology nowadays can be an active participant in customers’ decision-making (Pischetola et al., 2021 ). Stemming from agency theory (Jensen & Meckling, 1976 ), technology agency refers to the capacity of technology to collaborate and support human decision-making (Nyholm, 2020 ) and to perform independently and proactively on behalf of a human (Adams et al., 2022 ). It represents a new type of human-technology interaction, where technology is no longer a simple tool but an agentive actor (Morosan & Dursun-Cengizci, 2024 ). For example, in the hospitality industry, AI-based systems can assist in tasks such as room assignments, restaurant reservations, and other decision-making processes (Ivanov, 2023 ). The core advantage of such a system is its high autonomy and learning ability, allowing it to make decisions and act on behalf of the users without requiring human intervention; therefore, the technology agency level is high (Kennedy & Hidalgo, 2021 ). Similarly, how the level of technology agency in the restaurant ordering systems shapes diners’ intention to reduce food waste remains unclear. 2.3 Moral norms and intention to reduce food waste Intention is an individual’s willingness to exert effort towards a specific behaviour (Ajzen, 1991 ), and an individual’s intention is closely linked to his or her behaviour. Previous research has consistently shown that individuals’ intention is the strongest predictor of their behaviour, particularly within the context of reducing food waste (Teoh et al., 2022 ). Therefore, this study examines how diners’ IRFW is influenced by AI-driven services. Norms motivate individuals’ behaviour, especially when they are triggered by external factors (Han et al., 2018 ). Moral norms represent an individual’s perception of a moral obligation to comply with specific behaviours (Schwartz, 1977 ). Previous studies showed that individuals’ moral norms are essential drivers of individuals’ pro-environmental and sustainable behaviours, such as waste reduction and recycling behaviour (e.g. Han et al., 2018 ) and food waste decisions (e.g. Wang et al., 2021 ). Shan et al. ( 2024 ) explored the moderating effects of opportunity, ability, and face consciousness on the conversion of food waste reduction intention into actual behavior in the Chinese context. Bertoldo and Castro ( 2016 ) further noted that moral norms, as personal norms, are shaped by descriptive norms and injunctive norms (outer regulations). In this study, moral norms are defined as individuals’ beliefs about their moral obligation to reduce food waste (Talwar et al., 2022 ). 2.4 Hypotheses development As van Doorn et al. ( 2017 ) pointed out, service robots and other AI systems can evoke automated social presence that enables customers to perceive the presence of others, thereby activating their social cognitive mechanisms and influencing their behavioural tendencies. Kim et al.’s ( 2021 ) study on online education found that the social presence triggered by AI teachers can stimulate students’ social scripts, thereby fostering more positive attitudes and behavioural intentions. Lei, Xie et al.’s ( 2024 ) study on unethical consumer behaviour indicated that the high social presence created by partner roles or anthropomorphic designs in AI agents inhibits consumers’ anticipatory moral disengagement mechanisms, reducing acceptance of unethical behaviours. In this case, compared to technology with low social presence, technology (e.g., an AI-based ordering system) that allows customers to perceive high social presence, making diners feel that they are in the company of others, may activate their moral norms regarding the negative impact of food waste and reduce food waste. Therefore, it is proposed that: H1: The high social presence constructed by AI-assisted ordering systems enhances diners’ moral norms (a) and promotes diners’ IRFW (b) compared to traditional menu ordering systems (with low social presence). Brehmer’s ( 2023 ) study on psychology confirmed that individuals perceived moral norms mediate the relationship between cognitive attitudes, descriptive norms, perceived behavioural control (independent variables), and intention (dependent variable). Similarly, Han et al.’s ( 2018 ) study on sustainable tourism indicated that moral norms mediate the relationship between descriptive norms and the waste reduction and recycling intention of young travellers. In this case, AI-assisted order ingredients may trigger consumers’ moral norms, guiding them to take actual steps to reduce food waste. Therefore, diners’ moral norms may mediate the relationship between ordering systems and their IRFW. H2: Diners’ moral norms mediate the relationship between social presence constructed by ordering systems and IRFW. The human-computer interaction paradigm suggests that users unconsciously develop trust and compliance, similar to interpersonal relationships, towards highly intelligent and interactive technologies (Reeves & Nass, 1996 ; Nowak & Biocca, 2003 ). Human-technology integration aims to enhance the user’s capabilities, actions and optimise the outcomes (Patricia et al., 2022 ). High-level agency technology can empower users to make effective interactions and intelligent decisions, thereby effectively enhancing their ability to make sustainable choices. In this case, when an ordering system lacks social presence, high-level agency technology, such as AI-driven personalised recommendations and visual feedback on waste, can compensate by enhancing perceived expertise and credibility (Liew & Tan, 2021 ). This expert interaction can enhance users’ awareness of the consequences of their actions, thereby activating personal moral norms, as suggested by the norm activation model (Schwartz, 1977 ). On the other hand, systems that merely execute instructions and lack autonomy may not provide such persuasive cues, thus weakening the impact of social presence on moral awareness or behavioural change. Therefore, it is proposed that: H3: The technology agency level moderates the relationship between the social presence constructed by ordering systems and moral norms (a) and IRFW (b). More specifically, the high technology agency level strengthens the impact of social presence constructed by ordering systems on diners’ moral norms and IRFW. H4: Moral norms mediate the interaction effect of technology agency level and social presence constructed by ordering systems on diners’ IRFW. According to the above-mentioned hypotheses, a research model was proposed (Fig. 1 ). 3. Study 1 3.1 Experimental design 3.1.1 Stimuli Study 1 was a one-factor between-groups experimental design. A scenario-based design was used to investigate how social presence in the ordering system of a Chinese hot pot and barbecue buffet restaurant influences diners’ moral norms and IRFW. Scenario-based experiments are efficient and easy to implement, enabling a more comprehensive analysis of the relationships among stimuli, cognitive processes, and outcomes (Budovska et al., 2020 ). Two scenarios were developed, as shown in Table 1 . Table 1 Scenarios for Study 1 Low social presence in technology (Self-order) You are dining in a Chinese hot pot and barbecue buffet restaurant. You are free to choose the dishes and portion sizes you like, and you can place unlimited orders. The waiter just tells you that you just need to scan the QR code on the self-ordering system to order ingredients, and robots will deliver the dishes you order. You find that you can order freely. High social presence in technology (AI-assisted order) You are dining in a Chinese hotpot and grill buffet restaurant. You are free to choose the dishes and portion sizes you like, and you can place unlimited orders. The waiter just tells you that you just need to talk with the AI-assisted ordering system to order ingredients, and robots will deliver the dishes you order. You find that the AI-assisted ordering system talks to you like a human, and it is listening to your orders. For example, AI may chat with you, saying, “What is your favourite food?”, “This is a good choice”, and “Enjoy your meal”. Forty university students who had dining experience in Chinese hot pot and barbecue buffet restaurants were invited to conduct a manipulation test. They were randomly assigned to one of two scenarios. After viewing the scenario, they were invited to answer the question “I feel a sense of social presence constructed by this restaurant ordering system (feeling the presence of somebody).” on a 7-point Likert scale (1 = low sense of social presence, 7 = high sense of social presence). The mean value of the traditional self-order group was less than that of the AI-assisted order group (M self−order =1.619, M AI−assisted order =4.737, t-value=-11.761, p < 0.001). The difference between the two groups is significant, indicating successful manipulation. 3.1.2 Measurement items The questionnaire was structured into three sections. The initial segment consisted of three screening queries: “Do you have any reasons that you will not visit Chinese hot pot and barbecue buffet restaurants?”, “Have you dined at Chinese hot pot and barbecue buffet restaurants in the last three months?”, and “Are you above 18 years of age?” The first filter question is used to prevent individuals’ bias on Chinese hot pot and barbecue buffet restaurants from affecting the results. In addition, individuals over 18 years old without recent dining experiences at the restaurants qualified for the survey. Their recent dining experiences may create bias in answering the questionnaire. Also, a consent statement was included to explain their rights in participating in the experiment. The second segment encompassed items measuring moral norms and IRFW. To ensure reliability and validity, moral norms were evaluated using six items adapted from Talwar et al. ( 2022 ), which included statements such as: “Food wastage makes me feel guilty about the wastage of resources,” “Food wastage gives me a bad conscience,” “Food wastage is against my morals, ” “Food wastage makes me feel bad, ” “Food wastage gives me a feeling of regret, ” and “I feel ashamed if I waste food even if nobody is aware of my action.” The IRFW was measured by three items borrowed from Teoh et al. ( 2022 ). To fit the research context, the statements were revised as: “When ordering ingredients, I intend not to order too much to throw food away”, “When ordering ingredients, my goal is not to order too much to throw food”, and “When ordering ingredients, I will try not to order too much to throw food away”. All responses were recorded using a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). The final section of the questionnaire collected demographic details from respondents. The questionnaire was initially written in English and then professionally translated into Chinese. Subsequently, another professional translator back translated the Chinese version into English. Both versions were reviewed for content validity by two hospitality and tourism professors. A pilot test involving thirty university students confirmed the understanding of the questionnaire, leading to no further adjustments. 3.1.3 Sampling method Mall intercepts were conducted in three cities of the Guangdong-Hong Kong-Macao Greater Bay Area. Time-based systematic sampling was utilised to enhance the samples’ representativeness and eliminate the bias inherent in convenience sampling (Hair et al., 2019 ). The targets of the study are potential customers of Chinese hot pot and buffet restaurants. Research assistants randomly selected one large-scale shopping mall to conduct mall intercepts every day from 7 to 20 September 2025, between 10 a.m. and 8 p.m. Every 20 minutes, research assistants approached the closest potential respondent and explained the purpose of the survey. Upon confirmation of participation, three filter questions were asked to qualify respondents. Once respondents passed the filter questions, they were randomly allocated to read one scenario on an iPad. Following comprehension of the scenario, they were invited to fill out the questionnaire. 181 valid data were gathered for data analysis. Table 2 presents the demographic details of the respondents. Table 2 Demographic information for study 1 (n = 181) Frequency % Gender Male 80 44.2 Female 101 55.8 Age 18–20 21 11.6 21–30 94 51.9 31–40 48 26.5 41–50 15 8.3 > 50 3 1.7 Education High school or below 16 8.8 Diploma 62 34.3 University degree 87 48.1 Postgraduate 16 8.8 Condition Low social presence 91 50.3 High social presence 90 49.7 3.2 Research findings 3.2.2 Reliability test and common method bias test The Cronbach’s α values for moral norms and IRFW were 0.792 and 0.753, respectively, supporting strong reliability. 3.2.3 Main effect This study employed one-way ANCOVA to compare the differences in moral norms and IRFW among the two groups, while considering demographic characteristics such as gender, age, and education as covariates. The results showed a statistically significant effect of the social presence in technology on moral norms (F = 19.015, p < 0.001). As illustrated in Fig. 2 , the participants in the AI-assisted order condition exhibited higher moral norms (M = 5.876, S.D.=0.511) in comparison to participants in the self-order group (M = 5.460, S.D.=0.605), supporting H1a. Similarly, social presence had a significant influence on diners’ IRFW (F = 29.353, p < 0.001). Participants in the AI-assisted order condition exhibited a higher level of IRFW (M = 5.878, S.D.=0.459) than participants in the condition of self-order (M = 5.282, S.D.=0.744), supporting H1b. 3.2.4 Mediating effect The mediating effect was examined utilising the PROCESS model with 5000 subsamples and a 95% confidence interval (CI) in SmartPLS 4.0.8.9. Table 3 illustrates that moral norms exerted a significant direct influence on the IRFW (β = 0.324, CI=[0.126, 0.506]), and social presence in technology exerted a significant direct impact on moral norms (β = 0.416, CI=[0.259, 0.581]) as well as the IRFW (β = 0.461, CI= [0.263, 0.655]). Furthermore, social presence had a significant indirect effect on IRFW through moral norms (β = 0.135, CI=[0.041, 0.253]). These results suggested the mediating role of moral norms, supporting H2. Table 3 Study 1 - Mediation effect analysis Direct effect - confidence intervals Path Coefficient 2.50% 97.50% Moral norms ◊ intention to reduce food waste 0.324 0.126 0.506 Social presence ◊ Moral norms 0.461 0.263 0.655 Social presence ◊ intention to reduce food waste 0.416 0.259 0.581 Specific indirect effect - confidence intervals Path Coefficient 2.50% 97.50% Social presence ◊ Moral norms ◊ intention to reduce food waste 0.135 0.041 0.253 4 Study 2 4.1 Experimental design 4.1.2 Stimuli This study used a 2 social presence in technology (high vs. low) x 2 technology agency level (high vs. low) factorial between-groups experimental design. Similar to Study 1, a scenario-based design was used. Four scenarios were developed (Shown in Table 4 ). Table 4 Scenarios for Study 2 Low social presence, low technology agency level Same as Study 1 in the condition of “Low social presence in technology (Self-order)” High social presence, low technology agency level Same as Study 1 in the condition of “High social presence in technology (AI-assisted order)” Low social presence, high technology agency level You are dining in a Chinese hotpot and grill buffet restaurant. You are free to choose the dishes and portion sizes you like, and you can place unlimited orders. The waiter just tells you that you just need to scan the QR code on the self-ordering system to order ingredients, and robots will deliver the dishes you order. The self-ordering system provides you with dish suggestions based on the number of people at your table. You find that you can order freely. High social presence, high technology agency level You are dining in a Chinese hotpot and grill buffet restaurant. You are free to choose the dishes and portion sizes you like, and you can place unlimited orders. The waiter just tells you that you just need to talk with the AI-assisted ordering system to order ingredients, and robots will deliver the dishes you order. You find that the AI-assisted ordering system talks to you like a human, and it is listening to your orders. AI-assisted ordering system provides you with dish suggestions based on the number of people at your table. For example, AI may chat with you, saying, “What is your favorite food?” and “You three meal choices are appropriate; you can add more if they’re not enough. Enjoy your meal.” Sixty university students were invited to participate in a manipulation test. They were randomly assigned to one of four scenarios. After reading their respective scenarios, they were asked two questions: “I feel a sense of social presence constructed by this restaurant ordering system (feeling the presence of somebody)” on a 7-point Likert scale (1 = low sense of social presence, 7 = high sense of social presence) and “I feel this restaurant ordering system is actively suggesting me a lot in ordering ingredients” on a 7-point Likert scale (1 = low level of suggestions, 7 = high level of suggestions). The mean values of question 1 for the two traditional self-order groups were lower than the mean values for the AI-assisted order groups (M self-order =2.167, M AI-assisted order =4.833, t-value =-11.087, p < 0.001). The mean values of question 2 for the two no-dish suggestion groups were lower than the mean values for the dish suggestion groups (M no disk suggestion =2.567, M with dish suggestion =4.900, t-value =-8.736, p < 0.001). The differences between the four groups are significant, indicating successful manipulation. 4.1.2 Measurement items and sampling method This study employed the identical measurement items and the sampling method as those used in Study 1. Samples were assigned randomly to view a distinct scenario on an iPad. Following the reading of the scenario, participants were invited to fill out a questionnaire. From 5 to 25 October 2025, a total of 363 samples provided valid responses. The demographic details were reported in Table 5 . Table 5 Demographic information for Study 2 (n = 363) Frequency % Gender Male 175 48.2 Female 188 51.8 Age 18–20 57 15.7 21–30 142 39.1 31–40 110 30.3 41–50 44 12.1 > 50 10 2.8 Education High school or below 17 4.7 Diploma 106 29.2 University degree 168 46.3 Postgraduate 72 19.8 Condition Low social presence, low technology agency level 90 24.8 High social presence, low technology agency level 91 25.1 Low social presence, high technology agency level 90 24.8 High social presence, high technology agency level 92 25.3 4.2 Result 4.2.2 Reliability test The Cronbach’s α values for moral norms and IRFW were 0.741 and 0.780, respectively, supporting strong reliability. 4.2.3 Main effect, mediation effect, and moderation effect This study performed a two-way ANCOVA, with gender, age, and education as covariates, to compare the differences in moral norms and IRFW. The results indicated a statistically significant main effect of social presence (M self−order =5.459 vs. M AI−asssited order =5.942; F = 144.674, p < 0.001) and technology agency level (M no dish suggestion =5.255 vs. M with dish suggestion =6.147; F = 521.033, p < 0.001) on moral norms. There was also a significant effect of social presence (M self−order =5.533 vs. M AI−assisted order =5.994; F = 38.170, p < 0.001) and technology agency level (M no dish suggestion =5.567 vs. M with dish suggestion =5.963; F = 30.155, p < 0.001) on IRFW. So, H1 was supported. The results of the mediating effect (Table 6 ) showed that moral norms had a significant influence on the IRFW (β = 0.598, CI=[0.372, 0.799). Furthermore, social presence exerted a significant indirect influence on the IRFW through moral norms (β = 0.187, CI=[0.101, 0.278]). Hence, moral norms mediate the effect of social presence on IRFW, supporting H2. Table 6 Study 2 - Mediation effect analysis Direct effect-confidence intervals Path Coefficient 2.50% 97.50% Moral norms ◊ intention to reduce food waste 0.598 0.372 0.799 Social presence ◊ moral norms 0.312 0.204 0.425 Social presence ◊ intention to reduce food waste 0.448 0.266 0.628 Specific indirect effect-confidence intervals Path Coefficient 2.50% 97.50% Social presence ◊ moral norms◊ intention to reduce food waste 0.187 0.101 0.278 In addition, the results of two-way ANCOVA indicated that technology agency level significantly moderated the relationship between social presence and moral norms (F = 17.735, p < 0.001; Fig. 3 a) as well as the relationship between social presence and IRFW (F = 6.152, p = 0.014; Fig. 3 b). So, H3 were supported. Independent-sample t-tests were used to investigate the difference in moral norms between the self-order group and AI-assisted order group across two technology agency level conditions. In the low agency levels, the moral norms of the AI-assisted order group were significantly higher than those in the self-order group (M self−order =5.098 vs. M AI−assisted order =5.410; F = 0.302, p = 0.583, t=-5.610, p < 0.001). When agency levels were high, the moral norms in the AI-assisted order group were significantly higher than those in the self-order group too (M self−order =5.820 vs. M AI−assisted order =6.467; F = 3.362, p = 0.068, t=-11.489, p < 0.001). The results also showed that in the low agency levels, the IRFW in the AI-assisted ordering group was significantly higher compared to that in the self-order group (M self−order =5.248 vs. M AI−assisted order =5.883; F = 0.171, p = 0.680, t=-6.904, p < 0.001). In the high agency levels, the IRFW in the AI-assisted order group were significantly higher than those in the self-order group too (M self−order =5.819 vs. M AI−assisted order =6.105; F = 4.134, p = 0.043, t=-2.548, p = 0.012). 4.2.4 Ther result of moderated mediation effect The PROCESS procedure was performed to test the moderated mediation effect. Technology agency level was considered as the moderator for the relationship between social presence and IRFW, with moral norms acting as the mediator. When the technology agency level is low, the indirect effect of social presence on IRFW through moral norms was significant (β = 0.187, CI=[0.101, 0.278]). When the technology agency level is high, the indirect effect of social presence on IRFW through moral norms was significant too (β = 0.387, CI=[0.222, 0.565]). So, H4 was supported. 5. Discussion and implications 5.1 Discussion Study 1 validated that through AI-assisted order ingredients, diners perceive the presence of others while ordering ingredients that strengthen their moral norms, thereby reducing their intention to order less to reduce wasted food. AI-driven services place orders for diners in a human-like manner, creating social norms that convince diners that wasting food is not a common or acceptable practice in that environment. These findings align with existing literature supporting Lei, Xie et al. ( 2024 ) that social presence in human-technology interaction can shape users’ sustainable behaviour. Furthermore, the results confirmed the mediating role of moral norms. This result is consistent with the literature supporting the mediating role of personal norms (e.g., Han et al., 2018 ). Study 2 confirmed the interaction effects between social presence constructed by AI technology and technology agency level in AI-driven services on enhancing diners’ moral norms and the IRFW. The results suggest that social presence and technology agency level in AI-driven services can interact and complement each other, collectively prompting diners to reduce food waste. These results are consistent with the literature supporting that users trust and compliance towards intelligent systems (Liew & Tan, 2021 ). The interactive effect (Fig. 3 ) shows that the high level of technology agency can enhance consumers’ moral norms, leading to a heightened IRFW regardless of the low or high social presence constructed by AI technology. When perceiving higher social presence and level of technology agency, they collectively cultivate the highest level of moral norms, consequently yielding the highest food waste reduction intention. 5.2 Theoretical implication While significant research attention has been in reducing food waste, existing literature primarily focuses on supply chain management or inventory optimisation, such as predicting spoilage and optimising stock (e.g. Onyeaka et al., 2025 ). However, food waste extends beyond back-end kitchen waste to include consumer behavioural waste at the front-end, particularly in buffets due to their nature of “eat-all-you-can”. Therefore, researchers studied customers’ responses to reduce food waste based on the restaurant’s explicit information (Budovska et al., 2020 ; Chang, 2022 ). Responding to the rapid diffusion of adopting AI technology in different kinds of service industries, this study may be the first that explore how AI technologies help to create an environment that stimulates customers to make environmental responses in terms of reducing food waste in the restaurant context. This research contributes to environmental protection research in adopting AI technologies and calls for future research on how AI applications can interact with consumers to effectively change their behaviours more sustainably. This study uses the social presence of the AI-driven services and technology agency level in the food ordering process as stimuli. By investigating the intricate interplay between technology characteristics and internal personal standards (moral norms), this study provides insights to researchers to understand the complex dynamics of how advanced technology shapes human sustainable behaviours. In addition, by incorporating the norm activation model, the study reveals the underlying mechanisms through which the interaction of technological characteristics (social presence and technology agency level) activates users’ personal moral norms as an internal psychological process, ultimately influence the intention to reduce food waste. This study extends the moral theory from the traditional interpersonal social context to the field of human-technology interaction, providing a key theoretical explanation path for understanding how AI can become a moral promoter and shape users’ sustainable behaviour. This study shows an interesting finding that the technology agency level positively moderated the relationship between social presence in the ordering system and moral norms and negatively moderated the relationship between social presence in the ordering system and the IRFW (Fig. 3 ). It means that the high technology agency level strengthens the influence of social presence in the ordering system on consumers’ moral norms, thus influencing their IRFW. On the other hand, the high technology agency level supplements the low social presence in technology by inducing moral norms and enhancing consumers’ IRFW. This study enriches AI in service literature by confirming that the moderating role of technology agency level in shaping consumers’ responses can be in various ways. To create diners’ awareness of wasting food to reduce food waste, some buffet restaurants use punishment statements (Chang, 2022 ), and some restaurants use encouragement statements (Chang, 2022 ). These two methods are explicit approaches. This study tests the scenarios of using AI-driven services to enhance diners’ moral norms through the implication of advanced technology. It means that AI human-like features can introduce social elements and potentially leverage moral norms in the service process to develop customers’ social responsibility. Such implicit norm visualisation changes consumer decision-making environments in a non-confrontational manner, even proving more effective than traditional moral preaching methods (Schultz et al., 2007 ). It provides a novel perspective on inducing people’s moral norms to address food waste in the Chinese hot pot and barbecue buffet context. 5.3 Practical implications Restaurants (not only hot pot and barbecue buffet restaurants) are suggested to implement an AI strategy to eliminate food waste. AI should be trained to act like a waiter to interact with customers, which will enhance the perceived social presence while ordering. Leverages AI-driven services to enable real-time monitoring and provide real-time interaction to customers. For example, answer questions about the menu and offer information on specials during ordering. Furthermore, to enhance the social presence during human-technology interaction, it is suggested to design an AI waiter with humanity and warmth. Emojis can be used to enhance the interactive experience, making it easier for customers to engage with them pleasantly. When implementing AI-driven service strategies, training AI-assisted ordering systems is essential to effectively recommend portion sizes and menu choices to help customers make decisions. The AI-assisted ordering system developers should train the systems to be able to provide accurate suggestions for the number of dishes based on the number of diners. For example, telling customers that “for 4 people, 6 meat plates and 4 vegetable plates should be appropriate.” A reminder suggestion of reordering after finishing the current dishes is helpful. Since customers come from all around, AI-assisted ordering systems should be trained to understand customers’ preferences and dining habits, which helps to make dish suggestions. Additionally, AI can analyse customer feedback and behaviour to provide personalised suggestions on menu items, promoting a more efficient and tailored dining experience that minimises waste. For loyal customers, human-like AI-driven services can work as human waiters to suggest portion sizes based on their historical data and customer preferences, helping diners make more informed choices and avoid over-ordering. 5.4 Limitations Firstly, the samples were limited to the Guangdong-Hong Kong-Macao Greater Bay Area, China. Future studies may involve more diverse samples from other cities in China or samples from other countries. Secondly, this study only tested the Chinese hot pot and barbecue buffet restaurants. Future studies could take other types of restaurants. Thirdly, this study only considers the effect of social presence and the technology agency level. Future studies can take other characteristics of AI-based systems into consideration and investigate how these characteristics jointly affect customers’ sustainable behaviour. Lastly, this study only tested the use of AI-assisted ordering systems. Future studies can test other AI technologies in other hospitality and tourism settings. Declarations Competing interests The authors declare no competing interests. Ethical Approval This study has received ethical approval from XXXX (Approval No.: XXXX; Approval Date: XXXX). The questionnaire survey was fully anonymized, causing no harm to respondents, and did not involve sensitive personal information, commercial interests, or disclosure of respondents’ personal privacy. All procedures involving human participants complied with the ethical standards of the institutional research committee and the 1964 Declaration of Helsinki and its subsequent amendments or comparable ethical standards. Informed Consent We conducted questionnaire surveys from 07/09/2025 to 25/10/2025 among participants aged over 18 years with dining experience in XXXX, with no other special restrictions on gender, age, or education background. Informed consent was obtained from all participants prior to the survey. They were clearly informed of the research purpose, procedures, rights of voluntary participation, and the anonymization and confidential handling of data. Participants filled in the questionnaire only after confirming and agreeing to participate. Author Contribution Jing Zhang: Writing - review & editing, Writing - original draft,Formal analysis, Data curation, Conceptualization. Ivan Ka Wai Lai:Writing - review & editing, Writing - original draft, Supervision,Methodology, Conceptualization. Xiaohong Wu: Writing - review &editing, Writing-original draft，Supervision, Methodology,Conceptualization. Data Availability The datasets and questionnaire used in this study are available in the supplementary materials. All data and materials are fully accessible for peer review References Adams HS (2024) Marriott hotels’ AI technology leads to food waste reduction. Food & Drink. Retrieved from https://fooddigital.com/restaurants/marriott-hotels-ai-technology-leads-to-food-waste-reduction Adams J, Dedehayir O, O’Connor PA (2022) Theoretical model of technology, agency, and wellbeing, The XXXIII ISPIM Innovation Conference Innovating in a Digital World , Copenhagen, Denmark Ajzen I (1991) The theory of planned behavior. Organizational Behav Hum Decis Process 50(2):179–211 Alam SS, Kokash HA, Ahsan MN, Ahmed S (2025) Relationship between technology readiness, ai adoption and value creation in hospitality industry: moderating role of technological turbulence. Int J Hospitality Manage 127:104133 Anderson KC, Hansen H, Laverie D (2024) Technology has agency too! Disentangling technology's causal and structural agency in a service system. J Macromarketing 44(2):510–533 Bai S, Yu D, Han C, Yang M, Gupta BB, Arya V, Panigrahi PK, Tang R, He H, Zhao J (2024) Warmth trumps competence: The anthropomorphic interaction experience on intelligent service evaluation: Insights from the high-evoked automated social presence. Technol Forecast Soc Chang 204:123395 Bertoldo R, Castro P (2016) The outer influence inside us: Exploring the relation between social and personal norms. Resour Conserv Recycl 112:45–53 Bi C, Cui X, Sun Z, Jin Y (2025) Thinking AI or feeling AI? The effect of AI on consumers’ willingness to purchase healthy food from the perspective of nudge. Humanit Social Sci Commun 12(1):1032 Biocca F, Chad H, Burgoon JK (2003) Towards a more robust theory and measure of social presence: Review and suggested criteria. Presence 12(5):456–480 Bowen J, Morosan C (2018) Beware hospitality industry: The robots are coming. Worldw Hospitality Tourism Themes 10(6):726–733 Brehmer M (2023) Perceived moral norms in an extended theory of planned behavior in predicting university students’ bystander intentions toward relational bullying. Eur J Invest Health Psychol Educ 13(7):1202–1218 Budovska V, Delgado AT, Øgaard T (2020) Pro-environmental behaviour of hotel guests: application of the theory of planned behaviour and social norms to towel reuse. Tourism Hospitality Res 20(1):105–116 Casaló LV, Millastre-Valencia P, Belanche D, Flavián C (2025) Intelligence and humanness as key drivers of service value in Generative AI chatbots. Int J Hospitality Manage 128:104130 Chang YY (2022) All you can eat or all you can waste? Effects of alternate serving styles and inducements on food waste in buffet restaurants. Curr Issues Tourism 25(5):727–744 Che S, Liu C, Wang J, Wang J (2026) Can artificial intelligence drive enterprise green management innovation? A new perspective on harnessing intelligence. Technol Soc 85:103156 Clark QM, Kanavikar DB, Clark J, Donnelly PJ (2025) Exploring the potential of AI-driven food waste management strategies used in the hospitality industry for application in household settings. Front Artif Intell 7:1429477 Dhir A, Talwar S, Kaur P, Malibari A (2020) Food waste in hospitality and food services: A systematic literature review and framework development approach. J Clean Prod 270:122861 Filimonau V, Algboory H, Mohammed NK, Kadum H, Qasem JM, Muhialdin B (2023) Food waste and its management in the foodservice sector of a developing economy: An exploratory and preliminary study of a sample of restaurants in Iraq. Tourism Manage Perspective 45:101048 Hair JF, Risher JJ, Sarstedt M, Ringle CM (2019) When to use and how to report the results of PLS-SEM. Eur Bus Rev 31(1):2–24 Han H, Yu J, Kim HC, Kim W (2018) Impact of social/personal norms and willingness to sacrifice on young vacationers’ pro-environmental intentions for waste reduction and recycling. J Sustainable Tourism 26(12):2117–2133 Ivanov SH (2023) Automated decision-making. Foresight 25(1):4–19 Jensen MC, Meckling WH (1976) Theory of the firm: Managerial behavior, agency costs and ownership structure. J Financ Econ 3(4):305–360 Kennedy D, Hidalgo M (2021) Two-Way Human-Agent trust relationships in adaptive cognitive agent, adaptive tasking scenarios: Literature metadata analysis. In: Kurosu, M. (eds) Human-Computer Interaction. Theory, Methods and Tools. HCII 2021, LNCS 12762, pp 191–205. Springer, Cham Khan O (2018) Food Waste in China: Whose Fault Is It ? Retrieved from https://news.cgtn.com/news/3d3d514e7841544f7a457a6333566d54/share_p.html Kim J Jr., Xu KM, K., Sellnow DD (2021) I like my relational machine teacher: An AI instructor’s communication styles and social presence in online education. Int J Human–Computer Interact 37(18):1760–1770 Kim J, Merrill JK, Xu K, Kelly S (2022) Perceived credibility of an AI instructor in online education: The role of social presence and voice features. Comput Hum Behav 136:107383 Lei SS, Agyeiwaah E, Fong LHN, Choe JY (2024) Exploring food waste at a hospitality staff canteen with meteorological data. J Hospitality Tourism Res 48(4):607–621 Lei S, Xie L, Peng J (2024) Unethical consumer behavior following artificial intelligence agent encounters: The differential effect of AI agent roles and its boundary conditions. J Service Res 28(4):589–613 Liew TW, Tan SM (2021) Social cues and implications for designing expert and competent artificial agents: A systematic review. Telematics Inform 65:101721 Morosan C, Dursun-Cengizci A (2024) Letting AI make decisions for me: An empirical examination of hotel guests’ acceptance of technology agency. Int J Contemp Hospitality Manage 36(3):946–974 Narayanan A, Bharath E, Vijayakumar R (2024) Implications of virtual reality on environmental sustainability in restaurants based on AI . In 2024 10th International Conference on Communication and Signal Processing (ICCSP) (pp. 1488–1493). IEEE. https://doi.org/10.1109/ICCSP60870.2024.10544119 Nowak KL, Biocca F (2003) The effect of the agency and anthropomorphism on users’ sense of telepresence, copresence, and social presence in virtual environments. Presence: Teleoperators Virtual Environ 12(5):481–494 Nyholm S (2020) Humans and Robots: Ethics, Agency, and Anthropomorphism. Rowman and Littlefield Publishing Group, London Onyeaka H, Akinsemolu A, Miri T, Nnaji ND, Duan K, Pang G, Tamasiga P, Khalid S, Al-Sharify ZT, Ugwa C (2025) Artificial intelligence in food system: Innovative approach to minimizing food spoilage and food waste. J Agric Food Res 21:101895 Pang M, Zhang Q, Zhou J, Yin Q, Tan Q, Zhong X, Zhang Y, Zhao L, Yang Y, Hao Y, Wang C, Zhang P, Zhang L, Yang Y (2023) Dietary patterns and environmental impacts of Chongqing hotpot in China. Resour Conserv Recycling 198:107118 Patricia C, Patrick H, Kasper H, Orestis G, Joanna B, Sriram S, Marianna O (2022) The sense of agency in emerging technologies for human-computer integration: A review. Front NeuroSci 16:949138 Pischetola M, Thédiga de Miranda LV, Albuquerque P (2021) The invisible made visible through technologies’ agency: A sociomaterial inquiry on emergency remote teaching in higher education. Learn Media Technol 46(4):390–403 Reeves B, Nass CI (1996) The media equation: How people treat computers, television, and new media like real people and places. Center for the Study of Language and Information; Cambridge University Schwartz SH (1977) Normative influences on altruism. Adv Exp Soc Psychol 10:221–279 Schultz PW, Nolan JM, Cialdini RB, Goldstein NJ, Griskevicius V (2007) The constructive, destructive, and reconstructive power of social norms. Psychol Sci 18(5):429–434 Shan L, Lu Q, Tong X (2024) How to improve the consistency of consumers’ food waste reduction intentions and behaviors? An analysis based on the expanded Motivation–Opportunity–Ability framework. Humanit Social Sci Commun 11(1):1530 Stöckli S, Dorn M, Liechti S (2018) Normative prompts reduce consumer food waste in restaurants. Waste Manag 77:532–536 SWITCH-Asia (2024) Reducing Food Waste in China: Experiences from Six Cities. SWITCh-Asia. https://www.switch-asia.eu/news/reducing-food-waste-in-china-experiences-from-six-cities/ Szulecka J, Bradshaw C, Principato L (2024) Food waste governance architectures in Europe: Actors, steering modes, and harmonization trends. Global Challenges 8(11):2300265 Talwar S, Kaur P, Kumar S, Salo J, Dhir A (2022) The balancing act: how do moral norms and anticipated pride drive food waste/reduction behaviour? J Retailing Consumer Serv 66:102901 Teoh CW, Koay KY, Chai PS (2022) The role of social media in food waste prevention behaviour. Br Food J 124(5):1680–1696 van Doorn J, Mende M, Noble SM, Hulland J, Ostrom AL, Grewal D, Petersen JA (2017) Domo arigato Mr. Roboto: Emergence of automated social presence in organizational frontlines and customers’ service experiences. J Service Res 20(1):43–58 Wang L, Yang Y, Wang G (2022) The clean your plate campaign: Resisting table food waste in an unstable world. Sustainability 14:4699 Wang P, McCarthy B, Kapetanaki AB (2021) To be ethical or to be good? The impact of ‘Good Provider’ and moral norms on food waste decisions in two countries. Glob Environ Change 69:102300 World Population Review (2025) Food waste by country 2025. World Popul Rev https://worldpopulationreview.com/country-rankings/food-waste-by-country Zhao Y, Tao P, Zhang B, Huan C (2020) Contribution of Chinese hot pot and barbecue restaurants on indoor environmental parameters. Aerosol Air Qual Res 20:2925–2940 Additional Declarations No competing interests reported. Supplementary Files Questionnaire.docx study1181DATA.xlsx STUDY2363DATA.xlsx 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-8843012","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":610259219,"identity":"4aa896dc-310b-4ec6-bb6d-39124bca4757","order_by":0,"name":"Jing Zhang","email":"","orcid":"","institution":"City University of Macau","correspondingAuthor":false,"prefix":"","firstName":"Jing","middleName":"","lastName":"Zhang","suffix":""},{"id":610259220,"identity":"e1241214-36ae-4c22-9aca-5522aa76a3c8","order_by":1,"name":"Ivan Ka Wai Lai","email":"","orcid":"","institution":"Macao Polytechnic University","correspondingAuthor":false,"prefix":"","firstName":"Ivan","middleName":"Ka Wai","lastName":"Lai","suffix":""},{"id":610259221,"identity":"9756cddb-16f7-45cf-bbc4-bbdc0c95ff36","order_by":2,"name":"Xiaohong Wu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvElEQVRIiWNgGAWjYDACCSD+ACaBgIdYLYwzGCQkSNPCzMPAQIIWg9s9ZtI2NRZ1BjcSGB+8bWOI5iekRXLOGTPpnGMSEkAtzIZz2xhyZzYQ0MIvkbtNOrcBrIVNmheoZcMBAlrYQFosIVrYfxOlBWwLI9QWZqK0SM7I/2zZc0xCcuaZh82Sc85JEPaLwY20xBs/aur4+Y4nH/zwpswmt5+ADgRQOMAIMl6CaA0MDPKE3DMKRsEoGAUjFwAAiTM65/zEbuEAAAAASUVORK5CYII=","orcid":"","institution":"City University of Macau","correspondingAuthor":true,"prefix":"","firstName":"Xiaohong","middleName":"","lastName":"Wu","suffix":""}],"badges":[],"createdAt":"2026-02-10 16:08:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8843012/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8843012/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105330151,"identity":"26b508e3-2944-40ec-805f-fbc8809cbd43","added_by":"auto","created_at":"2026-03-24 20:35:41","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":228535,"visible":true,"origin":"","legend":"\u003cp\u003eProposed research model\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-8843012/v1/b4b696fbb02c7962b29d1a88.png"},{"id":105330152,"identity":"3516beac-0774-4ccd-b387-4d6f85948f3b","added_by":"auto","created_at":"2026-03-24 20:35:41","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":33468,"visible":true,"origin":"","legend":"\u003cp\u003eMoral norms and intention to reduce food waste\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8843012/v1/0b83760211521e15fa7e16ad.png"},{"id":105330153,"identity":"5019cb42-7c44-42e6-a771-db306a73f193","added_by":"auto","created_at":"2026-03-24 20:35:42","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":51466,"visible":true,"origin":"","legend":"\u003cp\u003eThe interactive effect of social presence in technology and technology agency level on moral norms and intention to reduce food waste\u003c/p\u003e\n\u003cp\u003ea Moral norms\u003c/p\u003e\n\u003cp\u003eb Intention to reduce food waste\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-8843012/v1/2c04489938bed59fd5d4b347.png"},{"id":108804208,"identity":"6d50b966-5bf8-4fe2-8655-9a535df3ce92","added_by":"auto","created_at":"2026-05-08 15:17:58","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":681709,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8843012/v1/2dc79aba-5543-4c17-8edd-f74b47c91842.pdf"},{"id":105565204,"identity":"224d47ab-c5c6-475b-9d66-6bf1fefc3f2e","added_by":"auto","created_at":"2026-03-27 12:52:23","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":13730,"visible":true,"origin":"","legend":"","description":"","filename":"Questionnaire.docx","url":"https://assets-eu.researchsquare.com/files/rs-8843012/v1/722234acba3cc9c5e3e4df1b.docx"},{"id":105330154,"identity":"b4a0975d-fd26-4710-9bd1-d543ce6d5025","added_by":"auto","created_at":"2026-03-24 20:35:42","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":20047,"visible":true,"origin":"","legend":"","description":"","filename":"study1181DATA.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8843012/v1/d9e9ef5bc2c7f050d9d1e647.xlsx"},{"id":105330155,"identity":"2ac00e72-f1b6-4060-b08b-6b7568326897","added_by":"auto","created_at":"2026-03-24 20:35:42","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":32850,"visible":true,"origin":"","legend":"","description":"","filename":"STUDY2363DATA.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8843012/v1/e5015c73c7ed7e34715fe8ca.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Using AI-driven services to reduce food waste in restaurants: The role of social presence and technology agency level","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eFood waste is a substantial environmental concern worldwide. The United Nations Sustainable Development Goals (UN SDGs) call for cutting global per capita food waste at the retail and consumer levels in half by 2030. Reducing food waste is a challenging task, a global focus, and has attracted great recent research interest (Filimonau et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Consequently, many studies on reducing food waste have been conducted, but most have been carried out in Western countries (Dhir et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In China, the catering industry is facing a significant food waste problem, with consumers throwing away between 17 and 18\u0026nbsp;million tons of food annually, an amount that could feed 30\u0026ndash;50\u0026nbsp;million people (Khan, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Despite the introduction of the \u0026ldquo;Clean Your Plate Campaign\u0026rdquo; (CYPC) in China over a decade ago, which heavily relies on personal ethics (Wang et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), progress remains limited, particularly in unique dining formats such as hot pot and barbecue buffets. These popular, self-cooking experiences, fundamentally different from Western pre-prepared buffets (Zhao et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), exacerbate waste through behaviours like over-selecting uncooked ingredients or leaving cooked food uneaten (Pang et al, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). As more and more Chinese hot pot and barbecue restaurants adopt self-ordering systems, customers can order uncooked ingredients by simply clicking on the menu. This quick operation makes it easier for customers to overorder. Given the popularity of self-ordering systems and the global spread of these restaurants, there is a need to investigate how to reduce consumers\u0026rsquo; food waste in hot pot and barbecue restaurants by using technologies, especially in China.\u003c/p\u003e \u003cp\u003eArtificial intelligence (AI) has been widely used in enterprise green management (Che et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2026\u003c/span\u003e). Leading businesses in the hospitality industry have begun leveraging AI to reduce food waste (Clark et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). For example, Marriott Hotel has successfully reduced food waste through the integration of the Winnow AI platform, which provides real-time insights into food consumption and waste, empowering the hotel to make smarter decisions around ordering, menu planning, and portion size (Adams, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Bi et al.\u0026rsquo;s (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) study on the impact of artificial intelligence on consumers' willingness to purchase healthy food and its role in reducing food waste. While current AI applications are primarily focused on back-end operations, AI can excel in front-end operations and even perform better in certain tasks. For instance, AI waiters in a restaurant can track customers\u0026rsquo; orders and provide personalised portion guidance like human staff. However, recent hospitality research on AI to reduce food waste focuses on its back-end functions (e.g. Adams, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Clark et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). There is a lack of empirical research on how AI-driven service models can promote sustainable dining behaviours, especially in high-waste environments like Chinese hot pot and barbecue buffets.\u003c/p\u003e \u003cp\u003eDifferent technologies may generate various senses of social presence, where technology is perceived as a sentient, social actor rather than a tool (Kim et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Recently, most restaurants adopted self-ordering systems, and Chinese hot pot and barbecue buffet restaurants are no exception. Self-ordering systems, typically scan QR codes, are functionally efficient but inherently low in social presence. When interacting with such a low social presence interface, customers do not have the sense of the presence of others; therefore, customers may choose a lot of ingredients without thinking carefully about whether to eat them all. However, AI-assisted ordering systems can communicate with customers just like a human waiter, which automatically evokes a sense of social presence (Bai et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Norms activation theory posits that individual behaviour is influenced by the activation of their personal norms (Schwartz, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e1977\u003c/span\u003e). Perceptions of AI as a human waiter can activate a sense of moral norms, preventing customers from ordering too much food and reducing waste.\u003c/p\u003e \u003cp\u003eIn addition, technology has different levels of agency, defined as its capability to act independently on behalf of the customer (Adams et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Systems with a low level of agency merely execute customer instructions, lacking effective guidance. In a restaurant setting, providing dish suggestions based on the number of diners and their preferences showcases a high level of agency. It shifts the technology from a passive tool to an active participant in diners\u0026rsquo; decision-making (Anderson et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), as they become active participants in shaping shared spatial experiences with humans, even shaping social norms (Pischetola et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Therefore, the technology agency level in providing disk suggestions may leverage and amplify the effect of social presence. Therefore, customers will have a higher sense of moral norms and will have a higher intention not to order too many dishes to reduce food waste when the system providing disk suggestions, whether using a self-ordering system or an AI-assisted ordering system.\u003c/p\u003e \u003cp\u003eThis study aims to investigate the ways of how AI-driven services can enhance consumers\u0026rsquo; moral norms to reduce food waste in Chinese hot pot and barbecue buffets. This study particularly answers below sub-questions: (1) Do the perceptions of social presence in the ordering system (high vs. low) influence consumers\u0026rsquo; moral norms and increase their IRFW? (2) Do moral norms mediate the relationship between ordering system and IRFW? and (3) Do the perceptions of technology agency level (high vs. low) moderate the effect of ordering system on customers\u0026rsquo; moral norms and IRFW? To answer the above sub-questions, this study employs a factorial experimental design and consists of two experiments. Study 1 is a one-factor between-groups experimental design, and Study 2 is a 2 x 2 factorial between-groups experimental design. Firstly, this study contributes to environmental protection research by adopting AI technologies to drive customers\u0026rsquo; environmental protection behaviours in the front-end food service context. Based on social presence theory, agency theory, and moral theory, this study explored how two types of features (constructing social presence and increasing technology agency level) of AI technologies implicitly induce moral norms, showing researchers and the catering business a new and friendly way to influence consumer sustainable behaviour. It also enriches the literature on AI-assisted services by studying how these two types of AI characteristics interact to influence customer responses. Finally, previous studies on reducing restaurant food waste mainly proposed legal approaches (Szulecka et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and campaign approaches (Wang et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and using explicit informational interventions as stimuli under human services to test the effect of social norms (e.g. Budovska et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). This study provides a novel perspective on inducing people\u0026rsquo;s moral norms to address food waste in China, especially in the Chinese hot pot and barbecue buffet context. In addition, practical suggestions are provided to catering businesses on using AI-driven services to reduce food waste.\u003c/p\u003e"},{"header":"2. Literature review","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Food waste and AI in food waste reduction\u003c/h2\u003e \u003cp\u003eFood waste refers to the edible portion of food that is discarded or not consumed. Food waste is an increasing issue (Lei, Agyeiwaah et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), leading to notable environmental, economic, and social consequences (St\u0026ouml;ckli et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Many countries have introduced anti-food waste laws to address food waste issues. For example, China legislated its first anti-food waste law (the Anti-food Waste Law of the People's Republic of China) in 2021 (SWITCH-Asia, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In addition, numerous campaigns and programs have emerged to reduce food waste, such as the \u0026ldquo;Love Food, Hate Waste\u0026rdquo; campaign in the UK and the CYPC in China. However, the effectiveness of most campaigns is short, including China\u0026rsquo;s CYPC. According to the report on food waste by country in 2025, the food waste per capita in 2024 in China was about 76kg, which was about a 19% increase from 2021 (64kg) (World Population Review, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eResearchers have explored a variety of strategies to reduce food waste, including using AI. AI is a technology that allows electronic devices to imitate human actions (Bowen \u0026amp; Morosan, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), which has the potential to revolutionise the restaurant industry by enhancing sustainable practices (Narayanan et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and reduce waste (Che et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2026\u003c/span\u003e). AI has been widely utilised in the hospitality and tourism industry and has attracted significant research interests, including employing AI to reduce food waste. For example, Onyeaka et al. (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) conducted a systematic review of AI technologies (e.g. machine learning, predictive analytics) for food spoilage prediction across supply chain stages, highlighting their potential to reduce waste while addressing data and ethical challenges in implementation. Narayanan et al. (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) proposed a holographic \u0026ldquo;AI+virtual reality (VR)\u0026rdquo; ordering system to enhance restaurant operational efficiency, reduce paper waste, and improve customer experience through digital menus, wireless orders, and intelligent recommendation algorithms, thereby promoting environmental sustainability in the catering industry. Despite research interest in AI in the hospitality sector, much of the existing literature is either conceptual (e.g. Narayanan et al, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) or literature review (e.g. Onyeaka et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), with empirical studies primarily centred on customers\u0026rsquo; adoption (e.g. Alam et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Consequently, there is a lack of research examining whether the utilisation of AI-driven services can lead customers to have an intention to reduce wasting food in restaurants compared to self-ordering services, and the underlying mechanisms.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Social presence and technology agency\u003c/h2\u003e \u003cp\u003eSocial presence is the sensation of being with others (Biocca et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Since people respond socially to both human and computer-controlled entities (Nowak \u0026amp; Biocca, 2023). With the technological advancements, the concept of social presence has evolved from human interaction to human-technology interaction (van Doorn et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Advanced technology, such as a robot, enables humans to experience automated social presence by creating a feeling of being in the presence of another social entity (van Doorn et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This feeling of social presence during technology interaction can be triggered by technological factors (Kim et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), such as humanisation (Casal\u0026oacute; et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). For example, the advancements in AI technology have the potential to evoke a strong sense of social presence during the interaction between customers and AI (Kim et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Social presence enhances positive mediated virtual experience (Biocca et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), playing a critical role in understanding how humans interact with machine agents (Kim et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, whether the role of social presence in the restaurant ordering systems can shape customers\u0026rsquo; intention to reduce food waste remains unclear.\u003c/p\u003e \u003cp\u003eTechnology nowadays can be an active participant in customers\u0026rsquo; decision-making (Pischetola et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Stemming from agency theory (Jensen \u0026amp; Meckling, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e1976\u003c/span\u003e), technology agency refers to the capacity of technology to collaborate and support human decision-making (Nyholm, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) and to perform independently and proactively on behalf of a human (Adams et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). It represents a new type of human-technology interaction, where technology is no longer a simple tool but an agentive actor (Morosan \u0026amp; Dursun-Cengizci, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). For example, in the hospitality industry, AI-based systems can assist in tasks such as room assignments, restaurant reservations, and other decision-making processes (Ivanov, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The core advantage of such a system is its high autonomy and learning ability, allowing it to make decisions and act on behalf of the users without requiring human intervention; therefore, the technology agency level is high (Kennedy \u0026amp; Hidalgo, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Similarly, how the level of technology agency in the restaurant ordering systems shapes diners\u0026rsquo; intention to reduce food waste remains unclear.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Moral norms and intention to reduce food waste\u003c/h2\u003e \u003cp\u003eIntention is an individual\u0026rsquo;s willingness to exert effort towards a specific behaviour (Ajzen, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e1991\u003c/span\u003e), and an individual\u0026rsquo;s intention is closely linked to his or her behaviour. Previous research has consistently shown that individuals\u0026rsquo; intention is the strongest predictor of their behaviour, particularly within the context of reducing food waste (Teoh et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Therefore, this study examines how diners\u0026rsquo; IRFW is influenced by AI-driven services.\u003c/p\u003e \u003cp\u003eNorms motivate individuals\u0026rsquo; behaviour, especially when they are triggered by external factors (Han et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Moral norms represent an individual\u0026rsquo;s perception of a moral obligation to comply with specific behaviours (Schwartz, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e1977\u003c/span\u003e). Previous studies showed that individuals\u0026rsquo; moral norms are essential drivers of individuals\u0026rsquo; pro-environmental and sustainable behaviours, such as waste reduction and recycling behaviour (e.g. Han et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) and food waste decisions (e.g. Wang et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Shan et al. (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) explored the moderating effects of opportunity, ability, and face consciousness on the conversion of food waste reduction intention into actual behavior in the Chinese context. Bertoldo and Castro (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) further noted that moral norms, as personal norms, are shaped by descriptive norms and injunctive norms (outer regulations). In this study, moral norms are defined as individuals\u0026rsquo; beliefs about their moral obligation to reduce food waste (Talwar et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Hypotheses development\u003c/h2\u003e \u003cp\u003eAs van Doorn et al. (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) pointed out, service robots and other AI systems can evoke automated social presence that enables customers to perceive the presence of others, thereby activating their social cognitive mechanisms and influencing their behavioural tendencies. Kim et al.\u0026rsquo;s (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) study on online education found that the social presence triggered by AI teachers can stimulate students\u0026rsquo; social scripts, thereby fostering more positive attitudes and behavioural intentions. Lei, Xie et al.\u0026rsquo;s (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) study on unethical consumer behaviour indicated that the high social presence created by partner roles or anthropomorphic designs in AI agents inhibits consumers\u0026rsquo; anticipatory moral disengagement mechanisms, reducing acceptance of unethical behaviours. In this case, compared to technology with low social presence, technology (e.g., an AI-based ordering system) that allows customers to perceive high social presence, making diners feel that they are in the company of others, may activate their moral norms regarding the negative impact of food waste and reduce food waste. Therefore, it is proposed that:\u003c/p\u003e \u003cp\u003eH1: The high social presence constructed by AI-assisted ordering systems enhances diners\u0026rsquo; moral norms (a) and promotes diners\u0026rsquo; IRFW (b) compared to traditional menu ordering systems (with low social presence).\u003c/p\u003e \u003cp\u003eBrehmer\u0026rsquo;s (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) study on psychology confirmed that individuals perceived moral norms mediate the relationship between cognitive attitudes, descriptive norms, perceived behavioural control (independent variables), and intention (dependent variable). Similarly, Han et al.\u0026rsquo;s (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) study on sustainable tourism indicated that moral norms mediate the relationship between descriptive norms and the waste reduction and recycling intention of young travellers. In this case, AI-assisted order ingredients may trigger consumers\u0026rsquo; moral norms, guiding them to take actual steps to reduce food waste. Therefore, diners\u0026rsquo; moral norms may mediate the relationship between ordering systems and their IRFW.\u003c/p\u003e \u003cp\u003eH2: Diners\u0026rsquo; moral norms mediate the relationship between social presence constructed by ordering systems and IRFW.\u003c/p\u003e \u003cp\u003eThe human-computer interaction paradigm suggests that users unconsciously develop trust and compliance, similar to interpersonal relationships, towards highly intelligent and interactive technologies (Reeves \u0026amp; Nass, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e1996\u003c/span\u003e; Nowak \u0026amp; Biocca, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Human-technology integration aims to enhance the user\u0026rsquo;s capabilities, actions and optimise the outcomes (Patricia et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). High-level agency technology can empower users to make effective interactions and intelligent decisions, thereby effectively enhancing their ability to make sustainable choices. In this case, when an ordering system lacks social presence, high-level agency technology, such as AI-driven personalised recommendations and visual feedback on waste, can compensate by enhancing perceived expertise and credibility (Liew \u0026amp; Tan, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). This expert interaction can enhance users\u0026rsquo; awareness of the consequences of their actions, thereby activating personal moral norms, as suggested by the norm activation model (Schwartz, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e1977\u003c/span\u003e). On the other hand, systems that merely execute instructions and lack autonomy may not provide such persuasive cues, thus weakening the impact of social presence on moral awareness or behavioural change. Therefore, it is proposed that:\u003c/p\u003e \u003cp\u003eH3: The technology agency level moderates the relationship between the social presence constructed by ordering systems and moral norms (a) and IRFW (b). More specifically, the high technology agency level strengthens the impact of social presence constructed by ordering systems on diners\u0026rsquo; moral norms and IRFW.\u003c/p\u003e \u003cp\u003eH4: Moral norms mediate the interaction effect of technology agency level and social presence constructed by ordering systems on diners\u0026rsquo; IRFW.\u003c/p\u003e \u003cp\u003eAccording to the above-mentioned hypotheses, a research model was proposed (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e\u0026lt;Please Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Study 1","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Experimental design\u003c/h2\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e3.1.1 Stimuli\u003c/h2\u003e \u003cp\u003eStudy 1 was a one-factor between-groups experimental design. A scenario-based design was used to investigate how social presence in the ordering system of a Chinese hot pot and barbecue buffet restaurant influences diners\u0026rsquo; moral norms and IRFW. Scenario-based experiments are efficient and easy to implement, enabling a more comprehensive analysis of the relationships among stimuli, cognitive processes, and outcomes (Budovska et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Two scenarios were developed, as shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\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\u003eScenarios for Study 1\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"1\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLow social presence in technology (Self-order)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYou are dining in a Chinese hot pot and barbecue buffet restaurant. You are free to choose the dishes and portion sizes you like, and you can place unlimited orders. The waiter just tells you that you just need to scan the QR code on the self-ordering system to order ingredients, and robots will deliver the dishes you order. You find that you can order freely.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh social presence in technology (AI-assisted order)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYou are dining in a Chinese hotpot and grill buffet restaurant. You are free to choose the dishes and portion sizes you like, and you can place unlimited orders. The waiter just tells you that you just need to talk with the AI-assisted ordering system to order ingredients, and robots will deliver the dishes you order. You find that the AI-assisted ordering system talks to you like a human, and it is listening to your orders. For example, AI may chat with you, saying, \u0026ldquo;What is your favourite food?\u0026rdquo;, \u0026ldquo;This is a good choice\u0026rdquo;, and \u0026ldquo;Enjoy your meal\u0026rdquo;.\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\u003e\u0026lt;Insert Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003cp\u003eForty university students who had dining experience in Chinese hot pot and barbecue buffet restaurants were invited to conduct a manipulation test. They were randomly assigned to one of two scenarios. After viewing the scenario, they were invited to answer the question \u0026ldquo;I feel a sense of social presence constructed by this restaurant ordering system (feeling the presence of somebody).\u0026rdquo; on a 7-point Likert scale (1\u0026thinsp;=\u0026thinsp;low sense of social presence, 7\u0026thinsp;=\u0026thinsp;high sense of social presence). The mean value of the traditional self-order group was less than that of the AI-assisted order group (M\u003csub\u003eself\u0026minus;order\u003c/sub\u003e=1.619, M\u003csub\u003eAI\u0026minus;assisted order\u003c/sub\u003e=4.737, t-value=-11.761, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The difference between the two groups is significant, indicating successful manipulation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003e3.1.2 Measurement items\u003c/h2\u003e \u003cp\u003eThe questionnaire was structured into three sections. The initial segment consisted of three screening queries: \u0026ldquo;Do you have any reasons that you will not visit Chinese hot pot and barbecue buffet restaurants?\u0026rdquo;, \u0026ldquo;Have you dined at Chinese hot pot and barbecue buffet restaurants in the last three months?\u0026rdquo;, and \u0026ldquo;Are you above 18 years of age?\u0026rdquo; The first filter question is used to prevent individuals\u0026rsquo; bias on Chinese hot pot and barbecue buffet restaurants from affecting the results. In addition, individuals over 18 years old without recent dining experiences at the restaurants qualified for the survey. Their recent dining experiences may create bias in answering the questionnaire. Also, a consent statement was included to explain their rights in participating in the experiment.\u003c/p\u003e \u003cp\u003eThe second segment encompassed items measuring moral norms and IRFW. To ensure reliability and validity, moral norms were evaluated using six items adapted from Talwar et al. (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), which included statements such as: \u0026ldquo;Food wastage makes me feel guilty about the wastage of resources,\u0026rdquo; \u0026ldquo;Food wastage gives me a bad conscience,\u0026rdquo; \u0026ldquo;Food wastage is against my morals, \u0026rdquo; \u0026ldquo;Food wastage makes me feel bad, \u0026rdquo; \u0026ldquo;Food wastage gives me a feeling of regret, \u0026rdquo; and \u0026ldquo;I feel ashamed if I waste food even if nobody is aware of my action.\u0026rdquo; The IRFW was measured by three items borrowed from Teoh et al. (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). To fit the research context, the statements were revised as: \u0026ldquo;When ordering ingredients, I intend not to order too much to throw food away\u0026rdquo;, \u0026ldquo;When ordering ingredients, my goal is not to order too much to throw food\u0026rdquo;, and \u0026ldquo;When ordering ingredients, I will try not to order too much to throw food away\u0026rdquo;. All responses were recorded using a 7-point Likert scale (1\u0026thinsp;=\u0026thinsp;strongly disagree, 7\u0026thinsp;=\u0026thinsp;strongly agree). The final section of the questionnaire collected demographic details from respondents.\u003c/p\u003e \u003cp\u003eThe questionnaire was initially written in English and then professionally translated into Chinese. Subsequently, another professional translator back translated the Chinese version into English. Both versions were reviewed for content validity by two hospitality and tourism professors. A pilot test involving thirty university students confirmed the understanding of the questionnaire, leading to no further adjustments.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e3.1.3 Sampling method\u003c/h2\u003e \u003cp\u003eMall intercepts were conducted in three cities of the Guangdong-Hong Kong-Macao Greater Bay Area. Time-based systematic sampling was utilised to enhance the samples\u0026rsquo; representativeness and eliminate the bias inherent in convenience sampling (Hair et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The targets of the study are potential customers of Chinese hot pot and buffet restaurants. Research assistants randomly selected one large-scale shopping mall to conduct mall intercepts every day from 7 to 20 September 2025, between 10 a.m. and 8 p.m. Every 20 minutes, research assistants approached the closest potential respondent and explained the purpose of the survey. Upon confirmation of participation, three filter questions were asked to qualify respondents. Once respondents passed the filter questions, they were randomly allocated to read one scenario on an iPad. Following comprehension of the scenario, they were invited to fill out the questionnaire. 181 valid data were gathered for data analysis. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the demographic details of the respondents.\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\u003eDemographic information for study 1 (n\u0026thinsp;=\u0026thinsp;181)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\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=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFrequency\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e44.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e55.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18\u0026ndash;20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21\u0026ndash;30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e51.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31\u0026ndash;40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e26.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e41\u0026ndash;50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eEducation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh school or below\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDiploma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e34.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUniversity degree\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e48.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePostgraduate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eCondition\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow social presence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e50.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh social presence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e49.7\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\u003e\u0026lt;Insert Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Research findings\u003c/h2\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e3.2.2 Reliability test and common method bias test\u003c/h2\u003e \u003cp\u003eThe Cronbach\u0026rsquo;s α values for moral norms and IRFW were 0.792 and 0.753, respectively, supporting strong reliability.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section3\"\u003e \u003ch2\u003e3.2.3 Main effect\u003c/h2\u003e \u003cp\u003eThis study employed one-way ANCOVA to compare the differences in moral norms and IRFW among the two groups, while considering demographic characteristics such as gender, age, and education as covariates. The results showed a statistically significant effect of the social presence in technology on moral norms (F\u0026thinsp;=\u0026thinsp;19.015, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, the participants in the AI-assisted order condition exhibited higher moral norms (M\u0026thinsp;=\u0026thinsp;5.876, S.D.=0.511) in comparison to participants in the self-order group (M\u0026thinsp;=\u0026thinsp;5.460, S.D.=0.605), supporting H1a.\u003c/p\u003e \u003cp\u003eSimilarly, social presence had a significant influence on diners\u0026rsquo; IRFW (F\u0026thinsp;=\u0026thinsp;29.353, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Participants in the AI-assisted order condition exhibited a higher level of IRFW (M\u0026thinsp;=\u0026thinsp;5.878, S.D.=0.459) than participants in the condition of self-order (M\u0026thinsp;=\u0026thinsp;5.282, S.D.=0.744), supporting H1b.\u003c/p\u003e \u003cp\u003e\u0026lt;Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section3\"\u003e \u003ch2\u003e3.2.4 Mediating effect\u003c/h2\u003e \u003cp\u003eThe mediating effect was examined utilising the PROCESS model with 5000 subsamples and a 95% confidence interval (CI) in SmartPLS 4.0.8.9. Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e illustrates that moral norms exerted a significant direct influence on the IRFW (β\u0026thinsp;=\u0026thinsp;0.324, CI=[0.126, 0.506]), and social presence in technology exerted a significant direct impact on moral norms (β\u0026thinsp;=\u0026thinsp;0.416, CI=[0.259, 0.581]) as well as the IRFW (β\u0026thinsp;=\u0026thinsp;0.461, CI= [0.263, 0.655]). Furthermore, social presence had a significant indirect effect on IRFW through moral norms (β\u0026thinsp;=\u0026thinsp;0.135, CI=[0.041, 0.253]). These results suggested the mediating role of moral norms, supporting H2.\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\u003eStudy 1 - Mediation effect analysis\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eDirect effect - confidence intervals\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePath\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.50%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e97.50%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMoral norms \u0026loz; intention to reduce food waste\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.324\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.126\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.506\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial presence \u0026loz; Moral norms\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.461\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.263\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.655\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial presence \u0026loz; intention to reduce food waste\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.416\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.259\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.581\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eSpecific indirect effect - confidence intervals\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePath\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.50%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e97.50%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial presence \u0026loz; Moral norms \u0026loz; intention to reduce food waste\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\u003e0.041\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.253\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\u003e\u0026lt;Insert Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"4 Study 2","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Experimental design\u003c/h2\u003e \u003cdiv id=\"Sec18\" class=\"Section3\"\u003e \u003ch2\u003e4.1.2 Stimuli\u003c/h2\u003e \u003cp\u003eThis study used a 2 social presence in technology (high vs. low) x 2 technology agency level (high vs. low) factorial between-groups experimental design. Similar to Study 1, a scenario-based design was used. Four scenarios were developed (Shown in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\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\u003eScenarios for Study 2\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"1\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLow social presence, low technology agency level\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSame as Study 1 in the condition of \u0026ldquo;Low social presence in technology (Self-order)\u0026rdquo;\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh social presence, low technology agency level\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSame as Study 1 in the condition of \u0026ldquo;High social presence in technology (AI-assisted order)\u0026rdquo;\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLow social presence, high technology agency level\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYou are dining in a Chinese hotpot and grill buffet restaurant. You are free to choose the dishes and portion sizes you like, and you can place unlimited orders. The waiter just tells you that you just need to scan the QR code on the self-ordering system to order ingredients, and robots will deliver the dishes you order. The self-ordering system provides you with dish suggestions based on the number of people at your table. You find that you can order freely.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh social presence, high technology agency level\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYou are dining in a Chinese hotpot and grill buffet restaurant. You are free to choose the dishes and portion sizes you like, and you can place unlimited orders. The waiter just tells you that you just need to talk with the AI-assisted ordering system to order ingredients, and robots will deliver the dishes you order. You find that the AI-assisted ordering system talks to you like a human, and it is listening to your orders. AI-assisted ordering system provides you with dish suggestions based on the number of people at your table. For example, AI may chat with you, saying, \u0026ldquo;What is your favorite food?\u0026rdquo; and \u0026ldquo;You three meal choices are appropriate; you can add more if they\u0026rsquo;re not enough. Enjoy your meal.\u0026rdquo;\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\u003e\u0026lt;Insert Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003cp\u003eSixty university students were invited to participate in a manipulation test. They were randomly assigned to one of four scenarios. After reading their respective scenarios, they were asked two questions: \u0026ldquo;I feel a sense of social presence constructed by this restaurant ordering system (feeling the presence of somebody)\u0026rdquo; on a 7-point Likert scale (1\u0026thinsp;=\u0026thinsp;low sense of social presence, 7\u0026thinsp;=\u0026thinsp;high sense of social presence) and \u0026ldquo;I feel this restaurant ordering system is actively suggesting me a lot in ordering ingredients\u0026rdquo; on a 7-point Likert scale (1\u0026thinsp;=\u0026thinsp;low level of suggestions, 7\u0026thinsp;=\u0026thinsp;high level of suggestions). The mean values of question 1 for the two traditional self-order groups were lower than the mean values for the AI-assisted order groups (M\u003csub\u003eself-order\u003c/sub\u003e=2.167, M\u003csub\u003eAI-assisted order\u003c/sub\u003e=4.833, t-value =-11.087, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The mean values of question 2 for the two no-dish suggestion groups were lower than the mean values for the dish suggestion groups (M\u003csub\u003eno disk suggestion\u003c/sub\u003e=2.567, M\u003csub\u003ewith dish suggestion\u003c/sub\u003e =4.900, t-value =-8.736, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The differences between the four groups are significant, indicating successful manipulation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section3\"\u003e \u003ch2\u003e4.1.2 Measurement items and sampling method\u003c/h2\u003e \u003cp\u003eThis study employed the identical measurement items and the sampling method as those used in Study 1. Samples were assigned randomly to view a distinct scenario on an iPad. Following the reading of the scenario, participants were invited to fill out a questionnaire. From 5 to 25 October 2025, a total of 363 samples provided valid responses. The demographic details were reported in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e.\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\u003eDemographic information for Study 2 (n\u0026thinsp;=\u0026thinsp;363)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\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=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFrequency\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eGender\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e175\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e48.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e188\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e51.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e \u003cp\u003e\u003cb\u003eAge\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18\u0026ndash;20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e15.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21\u0026ndash;30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e142\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e39.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31\u0026ndash;40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e41\u0026ndash;50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e12.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e\u003cb\u003eEducation\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh school or below\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDiploma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e106\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e29.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUniversity degree\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e168\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e46.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePostgraduate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e19.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e\u003cb\u003eCondition\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow social presence, low technology agency level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e24.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh social presence, low technology agency level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e25.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow social presence, high technology agency level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e24.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh social presence, high technology agency level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e25.3\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\u003e\u0026lt;Insert Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Result\u003c/h2\u003e \u003cdiv id=\"Sec21\" class=\"Section3\"\u003e \u003ch2\u003e4.2.2 Reliability test\u003c/h2\u003e \u003cp\u003eThe Cronbach\u0026rsquo;s α values for moral norms and IRFW were 0.741 and 0.780, respectively, supporting strong reliability.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section3\"\u003e \u003ch2\u003e4.2.3 Main effect, mediation effect, and moderation effect\u003c/h2\u003e \u003cp\u003eThis study performed a two-way ANCOVA, with gender, age, and education as covariates, to compare the differences in moral norms and IRFW. The results indicated a statistically significant main effect of social presence (M\u003csub\u003eself\u0026minus;order\u003c/sub\u003e=5.459 vs. M\u003csub\u003eAI\u0026minus;asssited order\u003c/sub\u003e=5.942; F\u0026thinsp;=\u0026thinsp;144.674, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and technology agency level (M\u003csub\u003eno dish suggestion\u003c/sub\u003e=5.255 vs. M\u003csub\u003ewith dish suggestion\u003c/sub\u003e=6.147; F\u0026thinsp;=\u0026thinsp;521.033, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) on moral norms. There was also a significant effect of social presence (M\u003csub\u003eself\u0026minus;order\u003c/sub\u003e=5.533 vs. M\u003csub\u003eAI\u0026minus;assisted order\u003c/sub\u003e=5.994; F\u0026thinsp;=\u0026thinsp;38.170, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and technology agency level (M\u003csub\u003eno dish suggestion\u003c/sub\u003e=5.567 vs. M\u003csub\u003ewith dish suggestion\u003c/sub\u003e=5.963; F\u0026thinsp;=\u0026thinsp;30.155, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) on IRFW. So, H1 was supported.\u003c/p\u003e \u003cp\u003eThe results of the mediating effect (Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e6\u003c/span\u003e) showed that moral norms had a significant influence on the IRFW (β\u0026thinsp;=\u0026thinsp;0.598, CI=[0.372, 0.799). Furthermore, social presence exerted a significant indirect influence on the IRFW through moral norms (β\u0026thinsp;=\u0026thinsp;0.187, CI=[0.101, 0.278]). Hence, moral norms mediate the effect of social presence on IRFW, supporting H2.\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\u003eStudy 2 - Mediation effect analysis\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eDirect effect-confidence intervals\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePath\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.50%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e97.50%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMoral norms \u0026loz; intention to reduce food waste\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.598\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.372\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.799\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial presence \u0026loz; moral norms\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.312\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.204\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.425\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial presence \u0026loz; intention to reduce food waste\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.448\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.266\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.628\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eSpecific indirect effect-confidence intervals\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePath\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.50%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e97.50%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSocial presence \u0026loz; moral norms\u0026loz; intention to reduce food waste\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.187\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.278\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\u003e\u0026lt;Insert Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e6\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003cp\u003eIn addition, the results of two-way ANCOVA indicated that technology agency level significantly moderated the relationship between social presence and moral norms (F\u0026thinsp;=\u0026thinsp;17.735, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea) as well as the relationship between social presence and IRFW (F\u0026thinsp;=\u0026thinsp;6.152, p\u0026thinsp;=\u0026thinsp;0.014; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). So, H3 were supported.\u003c/p\u003e \u003cp\u003e\u0026lt;Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u0026gt;\u003c/p\u003e \u003cp\u003eIndependent-sample t-tests were used to investigate the difference in moral norms between the self-order group and AI-assisted order group across two technology agency level conditions. In the low agency levels, the moral norms of the AI-assisted order group were significantly higher than those in the self-order group (M\u003csub\u003eself\u0026minus;order\u003c/sub\u003e=5.098 vs. M\u003csub\u003eAI\u0026minus;assisted order\u003c/sub\u003e=5.410; F\u0026thinsp;=\u0026thinsp;0.302, p\u0026thinsp;=\u0026thinsp;0.583, t=-5.610, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). When agency levels were high, the moral norms in the AI-assisted order group were significantly higher than those in the self-order group too (M\u003csub\u003eself\u0026minus;order\u003c/sub\u003e=5.820 vs. M\u003csub\u003eAI\u0026minus;assisted order\u003c/sub\u003e=6.467; F\u0026thinsp;=\u0026thinsp;3.362, p\u0026thinsp;=\u0026thinsp;0.068, t=-11.489, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003eThe results also showed that in the low agency levels, the IRFW in the AI-assisted ordering group was significantly higher compared to that in the self-order group (M\u003csub\u003eself\u0026minus;order\u003c/sub\u003e=5.248 vs. M\u003csub\u003eAI\u0026minus;assisted order\u003c/sub\u003e=5.883; F\u0026thinsp;=\u0026thinsp;0.171, p\u0026thinsp;=\u0026thinsp;0.680, t=-6.904, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). In the high agency levels, the IRFW in the AI-assisted order group were significantly higher than those in the self-order group too (M\u003csub\u003eself\u0026minus;order\u003c/sub\u003e=5.819 vs. M\u003csub\u003eAI\u0026minus;assisted order\u003c/sub\u003e=6.105; F\u0026thinsp;=\u0026thinsp;4.134, p\u0026thinsp;=\u0026thinsp;0.043, t=-2.548, p\u0026thinsp;=\u0026thinsp;0.012).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003e4.2.4 Ther result of moderated mediation effect\u003c/h2\u003e \u003cp\u003eThe PROCESS procedure was performed to test the moderated mediation effect. Technology agency level was considered as the moderator for the relationship between social presence and IRFW, with moral norms acting as the mediator. When the technology agency level is low, the indirect effect of social presence on IRFW through moral norms was significant (β\u0026thinsp;=\u0026thinsp;0.187, CI=[0.101, 0.278]). When the technology agency level is high, the indirect effect of social presence on IRFW through moral norms was significant too (β\u0026thinsp;=\u0026thinsp;0.387, CI=[0.222, 0.565]). So, H4 was supported.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"5. Discussion and implications","content":"\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Discussion\u003c/h2\u003e \u003cp\u003eStudy 1 validated that through AI-assisted order ingredients, diners perceive the presence of others while ordering ingredients that strengthen their moral norms, thereby reducing their intention to order less to reduce wasted food. AI-driven services place orders for diners in a human-like manner, creating social norms that convince diners that wasting food is not a common or acceptable practice in that environment. These findings align with existing literature supporting Lei, Xie et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) that social presence in human-technology interaction can shape users\u0026rsquo; sustainable behaviour. Furthermore, the results confirmed the mediating role of moral norms. This result is consistent with the literature supporting the mediating role of personal norms (e.g., Han et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eStudy 2 confirmed the interaction effects between social presence constructed by AI technology and technology agency level in AI-driven services on enhancing diners\u0026rsquo; moral norms and the IRFW. The results suggest that social presence and technology agency level in AI-driven services can interact and complement each other, collectively prompting diners to reduce food waste. These results are consistent with the literature supporting that users trust and compliance towards intelligent systems (Liew \u0026amp; Tan, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The interactive effect (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) shows that the high level of technology agency can enhance consumers\u0026rsquo; moral norms, leading to a heightened IRFW regardless of the low or high social presence constructed by AI technology. When perceiving higher social presence and level of technology agency, they collectively cultivate the highest level of moral norms, consequently yielding the highest food waste reduction intention.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section2\"\u003e \u003ch2\u003e5.2 Theoretical implication\u003c/h2\u003e \u003cp\u003eWhile significant research attention has been in reducing food waste, existing literature primarily focuses on supply chain management or inventory optimisation, such as predicting spoilage and optimising stock (e.g. Onyeaka et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, food waste extends beyond back-end kitchen waste to include consumer behavioural waste at the front-end, particularly in buffets due to their nature of \u0026ldquo;eat-all-you-can\u0026rdquo;. Therefore, researchers studied customers\u0026rsquo; responses to reduce food waste based on the restaurant\u0026rsquo;s explicit information (Budovska et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Chang, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Responding to the rapid diffusion of adopting AI technology in different kinds of service industries, this study may be the first that explore how AI technologies help to create an environment that stimulates customers to make environmental responses in terms of reducing food waste in the restaurant context. This research contributes to environmental protection research in adopting AI technologies and calls for future research on how AI applications can interact with consumers to effectively change their behaviours more sustainably.\u003c/p\u003e \u003cp\u003eThis study uses the social presence of the AI-driven services and technology agency level in the food ordering process as stimuli. By investigating the intricate interplay between technology characteristics and internal personal standards (moral norms), this study provides insights to researchers to understand the complex dynamics of how advanced technology shapes human sustainable behaviours. In addition, by incorporating the norm activation model, the study reveals the underlying mechanisms through which the interaction of technological characteristics (social presence and technology agency level) activates users\u0026rsquo; personal moral norms as an internal psychological process, ultimately influence the intention to reduce food waste. This study extends the moral theory from the traditional interpersonal social context to the field of human-technology interaction, providing a key theoretical explanation path for understanding how AI can become a moral promoter and shape users\u0026rsquo; sustainable behaviour.\u003c/p\u003e \u003cp\u003eThis study shows an interesting finding that the technology agency level positively moderated the relationship between social presence in the ordering system and moral norms and negatively moderated the relationship between social presence in the ordering system and the IRFW (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). It means that the high technology agency level strengthens the influence of social presence in the ordering system on consumers\u0026rsquo; moral norms, thus influencing their IRFW. On the other hand, the high technology agency level supplements the low social presence in technology by inducing moral norms and enhancing consumers\u0026rsquo; IRFW. This study enriches AI in service literature by confirming that the moderating role of technology agency level in shaping consumers\u0026rsquo; responses can be in various ways.\u003c/p\u003e \u003cp\u003eTo create diners\u0026rsquo; awareness of wasting food to reduce food waste, some buffet restaurants use punishment statements (Chang, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and some restaurants use encouragement statements (Chang, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). These two methods are explicit approaches. This study tests the scenarios of using AI-driven services to enhance diners\u0026rsquo; moral norms through the implication of advanced technology. It means that AI human-like features can introduce social elements and potentially leverage moral norms in the service process to develop customers\u0026rsquo; social responsibility. Such implicit norm visualisation changes consumer decision-making environments in a non-confrontational manner, even proving more effective than traditional moral preaching methods (Schultz et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). It provides a novel perspective on inducing people\u0026rsquo;s moral norms to address food waste in the Chinese hot pot and barbecue buffet context.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec27\" class=\"Section2\"\u003e \u003ch2\u003e5.3 Practical implications\u003c/h2\u003e \u003cp\u003eRestaurants (not only hot pot and barbecue buffet restaurants) are suggested to implement an AI strategy to eliminate food waste. AI should be trained to act like a waiter to interact with customers, which will enhance the perceived social presence while ordering. Leverages AI-driven services to enable real-time monitoring and provide real-time interaction to customers. For example, answer questions about the menu and offer information on specials during ordering. Furthermore, to enhance the social presence during human-technology interaction, it is suggested to design an AI waiter with humanity and warmth. Emojis can be used to enhance the interactive experience, making it easier for customers to engage with them pleasantly.\u003c/p\u003e \u003cp\u003eWhen implementing AI-driven service strategies, training AI-assisted ordering systems is essential to effectively recommend portion sizes and menu choices to help customers make decisions. The AI-assisted ordering system developers should train the systems to be able to provide accurate suggestions for the number of dishes based on the number of diners. For example, telling customers that \u0026ldquo;for 4 people, 6 meat plates and 4 vegetable plates should be appropriate.\u0026rdquo; A reminder suggestion of reordering after finishing the current dishes is helpful. Since customers come from all around, AI-assisted ordering systems should be trained to understand customers\u0026rsquo; preferences and dining habits, which helps to make dish suggestions. Additionally, AI can analyse customer feedback and behaviour to provide personalised suggestions on menu items, promoting a more efficient and tailored dining experience that minimises waste. For loyal customers, human-like AI-driven services can work as human waiters to suggest portion sizes based on their historical data and customer preferences, helping diners make more informed choices and avoid over-ordering.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003e5.4 Limitations\u003c/h2\u003e \u003cp\u003eFirstly, the samples were limited to the Guangdong-Hong Kong-Macao Greater Bay Area, China. Future studies may involve more diverse samples from other cities in China or samples from other countries. Secondly, this study only tested the Chinese hot pot and barbecue buffet restaurants. Future studies could take other types of restaurants. Thirdly, this study only considers the effect of social presence and the technology agency level. Future studies can take other characteristics of AI-based systems into consideration and investigate how these characteristics jointly affect customers\u0026rsquo; sustainable behaviour. Lastly, this study only tested the use of AI-assisted ordering systems. Future studies can test other AI technologies in other hospitality and tourism settings.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eEthical Approval\u003c/strong\u003e \u003cp\u003eThis study has received ethical approval from XXXX (Approval No.: XXXX; Approval Date: XXXX). The questionnaire survey was fully anonymized, causing no harm to respondents, and did not involve sensitive personal information, commercial interests, or disclosure of respondents\u0026rsquo; personal privacy. All procedures involving human participants complied with the ethical standards of the institutional research committee and the 1964 Declaration of Helsinki and its subsequent amendments or comparable ethical standards.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eInformed Consent\u003c/strong\u003e \u003cp\u003eWe conducted questionnaire surveys from 07/09/2025 to 25/10/2025 among participants aged over 18 years with dining experience in XXXX, with no other special restrictions on gender, age, or education background. Informed consent was obtained from all participants prior to the survey. They were clearly informed of the research purpose, procedures, rights of voluntary participation, and the anonymization and confidential handling of data. Participants filled in the questionnaire only after confirming and agreeing to participate.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eJing Zhang: Writing - review \u0026amp; editing, Writing - original draft,Formal analysis, Data curation, Conceptualization. Ivan Ka Wai Lai:Writing - review \u0026amp; editing, Writing - original draft, Supervision,Methodology, Conceptualization. Xiaohong Wu: Writing - review \u0026amp;editing, Writing-original draft，Supervision, Methodology,Conceptualization.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets and questionnaire used in this study are available in the supplementary materials. All data and materials are fully accessible for peer review\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAdams HS (2024) \u003cem\u003eMarriott hotels\u0026rsquo; AI technology leads to food waste reduction. Food \u0026amp; Drink.\u003c/em\u003e Retrieved from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://fooddigital.com/restaurants/marriott-hotels-ai-technology-leads-to-food-waste-reduction\u003c/span\u003e\u003cspan address=\"https://fooddigital.com/restaurants/marriott-hotels-ai-technology-leads-to-food-waste-reduction\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAdams J, Dedehayir O, O\u0026rsquo;Connor PA (2022) Theoretical model of technology, agency, and wellbeing, The XXXIII ISPIM Innovation Conference \u003cem\u003eInnovating in a Digital World\u003c/em\u003e, Copenhagen, Denmark\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAjzen I (1991) The theory of planned behavior. Organizational Behav Hum Decis Process 50(2):179\u0026ndash;211\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlam SS, Kokash HA, Ahsan MN, Ahmed S (2025) Relationship between technology readiness, ai adoption and value creation in hospitality industry: moderating role of technological turbulence. Int J Hospitality Manage 127:104133\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnderson KC, Hansen H, Laverie D (2024) Technology has agency too! Disentangling technology's causal and structural agency in a service system. J Macromarketing 44(2):510\u0026ndash;533\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBai S, Yu D, Han C, Yang M, Gupta BB, Arya V, Panigrahi PK, Tang R, He H, Zhao J (2024) Warmth trumps competence: The anthropomorphic interaction experience on intelligent service evaluation: Insights from the high-evoked automated social presence. Technol Forecast Soc Chang 204:123395\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBertoldo R, Castro P (2016) The outer influence inside us: Exploring the relation between social and personal norms. Resour Conserv Recycl 112:45\u0026ndash;53\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBi C, Cui X, Sun Z, Jin Y (2025) Thinking AI or feeling AI? The effect of AI on consumers\u0026rsquo; willingness to purchase healthy food from the perspective of nudge. Humanit Social Sci Commun 12(1):1032\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBiocca F, Chad H, Burgoon JK (2003) Towards a more robust theory and measure of social presence: Review and suggested criteria. Presence 12(5):456\u0026ndash;480\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBowen J, Morosan C (2018) Beware hospitality industry: The robots are coming. Worldw Hospitality Tourism Themes 10(6):726\u0026ndash;733\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrehmer M (2023) Perceived moral norms in an extended theory of planned behavior in predicting university students\u0026rsquo; bystander intentions toward relational bullying. Eur J Invest Health Psychol Educ 13(7):1202\u0026ndash;1218\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBudovska V, Delgado AT, \u0026Oslash;gaard T (2020) Pro-environmental behaviour of hotel guests: application of the theory of planned behaviour and social norms to towel reuse. Tourism Hospitality Res 20(1):105\u0026ndash;116\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCasal\u0026oacute; LV, Millastre-Valencia P, Belanche D, Flavi\u0026aacute;n C (2025) Intelligence and humanness as key drivers of service value in Generative AI chatbots. Int J Hospitality Manage 128:104130\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChang YY (2022) All you can eat or all you can waste? Effects of alternate serving styles and inducements on food waste in buffet restaurants. Curr Issues Tourism 25(5):727\u0026ndash;744\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChe S, Liu C, Wang J, Wang J (2026) Can artificial intelligence drive enterprise green management innovation? A new perspective on harnessing intelligence. Technol Soc 85:103156\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClark QM, Kanavikar DB, Clark J, Donnelly PJ (2025) Exploring the potential of AI-driven food waste management strategies used in the hospitality industry for application in household settings. Front Artif Intell 7:1429477\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDhir A, Talwar S, Kaur P, Malibari A (2020) Food waste in hospitality and food services: A systematic literature review and framework development approach. J Clean Prod 270:122861\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFilimonau V, Algboory H, Mohammed NK, Kadum H, Qasem JM, Muhialdin B (2023) Food waste and its management in the foodservice sector of a developing economy: An exploratory and preliminary study of a sample of restaurants in Iraq. Tourism Manage Perspective 45:101048\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHair JF, Risher JJ, Sarstedt M, Ringle CM (2019) When to use and how to report the results of PLS-SEM. Eur Bus Rev 31(1):2\u0026ndash;24\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan H, Yu J, Kim HC, Kim W (2018) Impact of social/personal norms and willingness to sacrifice on young vacationers\u0026rsquo; pro-environmental intentions for waste reduction and recycling. J Sustainable Tourism 26(12):2117\u0026ndash;2133\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIvanov SH (2023) Automated decision-making. Foresight 25(1):4\u0026ndash;19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJensen MC, Meckling WH (1976) Theory of the firm: Managerial behavior, agency costs and ownership structure. J Financ Econ 3(4):305\u0026ndash;360\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKennedy D, Hidalgo M (2021) Two-Way Human-Agent trust relationships in adaptive cognitive agent, adaptive tasking scenarios: Literature metadata analysis. \u003cem\u003eIn: Kurosu, M. (eds) Human-Computer Interaction. Theory, Methods and Tools.\u003c/em\u003e HCII 2021, LNCS 12762, pp 191\u0026ndash;205. Springer, Cham\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan O (2018) \u003cem\u003eFood Waste in China: Whose Fault Is It\u003c/em\u003e? Retrieved from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://news.cgtn.com/news/3d3d514e7841544f7a457a6333566d54/share_p.html\u003c/span\u003e\u003cspan address=\"https://news.cgtn.com/news/3d3d514e7841544f7a457a6333566d54/share_p.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim J Jr., Xu KM, K., Sellnow DD (2021) I like my relational machine teacher: An AI instructor\u0026rsquo;s communication styles and social presence in online education. Int J Human\u0026ndash;Computer Interact 37(18):1760\u0026ndash;1770\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim J, Merrill JK, Xu K, Kelly S (2022) Perceived credibility of an AI instructor in online education: The role of social presence and voice features. Comput Hum Behav 136:107383\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLei SS, Agyeiwaah E, Fong LHN, Choe JY (2024) Exploring food waste at a hospitality staff canteen with meteorological data. J Hospitality Tourism Res 48(4):607\u0026ndash;621\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLei S, Xie L, Peng J (2024) Unethical consumer behavior following artificial intelligence agent encounters: The differential effect of AI agent roles and its boundary conditions. J Service Res 28(4):589\u0026ndash;613\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiew TW, Tan SM (2021) Social cues and implications for designing expert and competent artificial agents: A systematic review. Telematics Inform 65:101721\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorosan C, Dursun-Cengizci A (2024) Letting AI make decisions for me: An empirical examination of hotel guests\u0026rsquo; acceptance of technology agency. Int J Contemp Hospitality Manage 36(3):946\u0026ndash;974\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNarayanan A, Bharath E, Vijayakumar R (2024) \u003cem\u003eImplications of virtual reality on environmental sustainability in restaurants based on AI\u003c/em\u003e. In 2024 10th International Conference on Communication and Signal Processing (ICCSP) (pp. 1488\u0026ndash;1493). IEEE. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/ICCSP60870.2024.10544119\u003c/span\u003e\u003cspan address=\"10.1109/ICCSP60870.2024.10544119\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNowak KL, Biocca F (2003) The effect of the agency and anthropomorphism on users\u0026rsquo; sense of telepresence, copresence, and social presence in virtual environments. Presence: Teleoperators Virtual Environ 12(5):481\u0026ndash;494\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNyholm S (2020) Humans and Robots: Ethics, Agency, and Anthropomorphism. Rowman and Littlefield Publishing Group, London\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOnyeaka H, Akinsemolu A, Miri T, Nnaji ND, Duan K, Pang G, Tamasiga P, Khalid S, Al-Sharify ZT, Ugwa C (2025) Artificial intelligence in food system: Innovative approach to minimizing food spoilage and food waste. J Agric Food Res 21:101895\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePang M, Zhang Q, Zhou J, Yin Q, Tan Q, Zhong X, Zhang Y, Zhao L, Yang Y, Hao Y, Wang C, Zhang P, Zhang L, Yang Y (2023) Dietary patterns and environmental impacts of Chongqing hotpot in China. Resour Conserv Recycling 198:107118\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatricia C, Patrick H, Kasper H, Orestis G, Joanna B, Sriram S, Marianna O (2022) The sense of agency in emerging technologies for human-computer integration: A review. Front NeuroSci 16:949138\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePischetola M, Th\u0026eacute;diga de Miranda LV, Albuquerque P (2021) The invisible made visible through technologies\u0026rsquo; agency: A sociomaterial inquiry on emergency remote teaching in higher education. Learn Media Technol 46(4):390\u0026ndash;403\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReeves B, Nass CI (1996) The media equation: How people treat computers, television, and new media like real people and places. Center for the Study of Language and Information; Cambridge University\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchwartz SH (1977) Normative influences on altruism. Adv Exp Soc Psychol 10:221\u0026ndash;279\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchultz PW, Nolan JM, Cialdini RB, Goldstein NJ, Griskevicius V (2007) The constructive, destructive, and reconstructive power of social norms. Psychol Sci 18(5):429\u0026ndash;434\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShan L, Lu Q, Tong X (2024) How to improve the consistency of consumers\u0026rsquo; food waste reduction intentions and behaviors? An analysis based on the expanded Motivation\u0026ndash;Opportunity\u0026ndash;Ability framework. Humanit Social Sci Commun 11(1):1530\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSt\u0026ouml;ckli S, Dorn M, Liechti S (2018) Normative prompts reduce consumer food waste in restaurants. Waste Manag 77:532\u0026ndash;536\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSWITCH-Asia (2024) Reducing Food Waste in China: Experiences from Six Cities. SWITCh-Asia. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.switch-asia.eu/news/reducing-food-waste-in-china-experiences-from-six-cities/\u003c/span\u003e\u003cspan address=\"https://www.switch-asia.eu/news/reducing-food-waste-in-china-experiences-from-six-cities/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSzulecka J, Bradshaw C, Principato L (2024) Food waste governance architectures in Europe: Actors, steering modes, and harmonization trends. Global Challenges 8(11):2300265\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTalwar S, Kaur P, Kumar S, Salo J, Dhir A (2022) The balancing act: how do moral norms and anticipated pride drive food waste/reduction behaviour? J Retailing Consumer Serv 66:102901\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTeoh CW, Koay KY, Chai PS (2022) The role of social media in food waste prevention behaviour. Br Food J 124(5):1680\u0026ndash;1696\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Doorn J, Mende M, Noble SM, Hulland J, Ostrom AL, Grewal D, Petersen JA (2017) Domo arigato Mr. Roboto: Emergence of automated social presence in organizational frontlines and customers\u0026rsquo; service experiences. J Service Res 20(1):43\u0026ndash;58\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang L, Yang Y, Wang G (2022) The clean your plate campaign: Resisting table food waste in an unstable world. Sustainability 14:4699\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang P, McCarthy B, Kapetanaki AB (2021) To be ethical or to be good? The impact of \u0026lsquo;Good Provider\u0026rsquo; and moral norms on food waste decisions in two countries. Glob Environ Change 69:102300\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWorld Population Review (2025) Food waste by country 2025. World Popul Rev \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://worldpopulationreview.com/country-rankings/food-waste-by-country\u003c/span\u003e\u003cspan address=\"https://worldpopulationreview.com/country-rankings/food-waste-by-country\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao Y, Tao P, Zhang B, Huan C (2020) Contribution of Chinese hot pot and barbecue restaurants on indoor environmental parameters. Aerosol Air Qual Res 20:2925\u0026ndash;2940\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":"AI-driven services, social presence, technology agency, food waste reduction, moral norms","lastPublishedDoi":"10.21203/rs.3.rs-8843012/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8843012/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study investigates how AI-driven services can enhance consumers\u0026rsquo; moral norms to reduce food waste in Chinese hot pot and barbecue buffets. A between-subjects 2 social presence in ordering system (high vs. low) x 2 technology agency level (high vs. low) factorial experimental design was used. Two experiments were performed via face-to-face questionnaire surveys with 181 and 363 respondents, respectively. The results indicate that high social presence (via AI-assisted ordering) enhances consumers\u0026rsquo; moral norms and their intention to reduce food waste (IRFW). The high agency level (providing dish suggestions) moderates the effect of ordering system on moral norms and IRFW, and the perception of moral norms mediates the interactive effects of social presence and technology agency level on IRFW. This study contributes to the field of hospitality technology research by examining the role of AI-driven services in reducing food waste in restaurants.\u003c/p\u003e","manuscriptTitle":"Using AI-driven services to reduce food waste in restaurants: The role of social presence and technology agency level","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-24 20:35:37","doi":"10.21203/rs.3.rs-8843012/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":"2e6bd540-0199-453a-b13d-5f143ccb0693","owner":[],"postedDate":"March 24th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Rejected","date":"2026-05-04T22:34:32+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":64929645,"name":"Business and commerce/Business and management"},{"id":64929646,"name":"Social science/Business and management"},{"id":64929647,"name":"Earth and environmental sciences/Environmental social sciences"},{"id":64929648,"name":"Biological sciences/Psychology"},{"id":64929649,"name":"Social science/Psychology"}],"tags":[],"updatedAt":"2026-05-04T22:39:13+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-24 20:35:37","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8843012","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8843012","identity":"rs-8843012","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}