Trusting AI Too Much? Psychological Predictors of Overtrust and the Mitigating Role of AI Literacy

Trusting AI Too Much? Psychological Predictors of Overtrust and the Mitigating Role of AI Literacy | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Trusting AI Too Much? Psychological Predictors of Overtrust and the Mitigating Role of AI Literacy Youngsang Kim, Jang Hyun Kim, Shunan Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7315296/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 11 You are reading this latest preprint version Abstract Background The increasing adoption of AI tools such as ChatGPT in academic contexts has introduced a new psychological risk: overtrust in AI. However, the psychological foundations and boundary conditions of AI overtrust remain insufficiently examined. This study examined how self-efficacy, intrinsic motivation, and extrinsic motivation influence AI overtrust, and whether AI literacy can mitigate its negative consequences. Methods A cross-sectional survey was conducted with 300 university students in South Korea who had prior academic experience using AI tools. The questionnaire measured academic self-efficacy, intrinsic and extrinsic motivation, AI literacy, and AI overtrust. Structural equation modeling was used to test direct and mediated effects among the variables, with bootstrapping employed to evaluate indirect pathways. Results Self-efficacy and intrinsic motivation were positively associated with AI overtrust, suggesting that learners with higher confidence and autonomy exhibit overtrust in AI-generated information. Extrinsic motivation showed a weaker direct effect. AI literacy had a significant negative effect on AI overtrust and mediated the relationship between both self-efficacy and intrinsic motivation and overtrust. A sequential mediation model indicated that performance expectation and AI literacy together suppressed overtrust by strengthening reflective awareness. Conclusions These findings demonstrate that high-performing and intrinsically motivated learners are psychologically vulnerable to AI overtrust. Although high academic self-efficacy fosters autonomy in using AI tools, it also increases the risk of overtrust in AI. Targeted AI literacy education addresses this risk by enhancing learners’ critical thinking, ethical awareness, and responsible AI use. AI literacy education should be adapted to learners’ psychological profiles to effectively prevent AI overtrust and support responsible use. AI overtrust self-efficacy intrinsic motivation extrinsic motivation AI literacy performance expectation structural equation modeling Figures Figure 1 Figure 2 Figure 3 1. Introduction As of the 2025, the advancement of artificial intelligence (AI) has brought about remarkable changes in the field of education, with generative AI such as ChatGPT leading innovative transformations in this area (Bhullar et al., 2024 ; Jo, 2024 ). ChatGPT, a generative AI chatbot developed by OpenAI, utilizes natural language processing technology and large-scale language models (LLMs) to provide human-like conversational experiences (Bettayeb et al., 2024 ). ChatGPT is widely used by university students around the world for various purposes, including creative content generation, question answering, and language translation (Jo, 2024 ; Li et al., 2024 ; Alzain et al., 2024 ). ChatGPT offers various benefits in the educational environment, such as personalized learning support and the promotion of critical thinking (Agbong-Coates, 2024 ; Mai et al., 2024 ). Students can receive assistance in their learning processes through close interaction with ChatGPT, and they can also receive real-time feedback from ChatGPT as a virtual teacher (Sok & Heng, 2023 ). These features encourage student engagement and ensure the self-directed learning (Bhullar et al., 2024 ). However, there are several threats to the successful integration of ChatGPT in education (Hasanein & Sobaih, 2023 ; Kovari, 2025 ). Data privacy concerns, students’ AI anxiety, fairness of academic outcomes derived from ChatGPT, and overtrust in AI are obstacles that can hinder successful learning (Bettayeb et al., 2024 ; Sok & Heng, 2023 ; Jo, 2024 ). In particular, the phenomenon of AI overtrust, known as the AI trust paradox, refers to excessive trust on AI beyond its actual capabilities (Ullrich et al., 2021 ). It can cause direct difficulties in the learning and teaching processes for both students and teachers (Pithers & Soden, 2000 ). Overtrust in AI can be understood as a cognitive bias, where users exhibit misplaced confidence in AI outputs despite uncertain validity (Li et al., 2024 ). It can result not only in poor decision-making and decreased performance but also even in serious safety incidents (Parasuraman & Manzey, 2010). Moreover, overtrust in AI diminishes human critical engagement, thereby negatively impacting both collaborative outcomes with AI and ethical accountability (Ullrich et al., 2021 ). Thus, overtrust in AI constitutes a significant risk factor at both individual and societal levels. It is essential for learners to develop AI literacy in the educational field, as it helps them integrate critical thinking and develop an ethical awareness of AI (Laitinen et al., 2017 ). Fostering AI literacy in students is crucial to addressing significant AI-related side effects, such as overtrust in AI (Mohebi, 2024 ). While prior studies have consistently emphasized the necessity of enhancing ethical and analytical use of AI through AI literacy, empirical research examining the practical role of AI literacy in addressing of AI overtrust remains limited. In particular, although existing research has primarily focused on the risks of general machine overtrust (Lee & See, 2004 ; Kaur et al., 2022 ; Ullrich et al., 2021 ), psychological research that identifies the antecedents of AI overtrust remains scarce. While there is some research on the psychological factors influencing AI dependency (Ye et al., 2025 ; Acosta-Enriquez et al., 2025 ; Zhang et al., 2024 ), little attention has been paid to investigating the psychological antecedents of AI overtrust. Therefore, this study contributes to understanding the psychological mechanisms underlying AI overtrust and examines whether AI literacy functions as a cognitive and behavioral buffer against its adverse consequences. 2. Literature review 2.1 AI overtrust and Learner Traits: Self-Efficacy and Motivation AI overtrust is considered a psychological state in which an individual’s level of trust in AI exceeds its actual capabilities (Aroyo et al., 2021 ; Ullrich et al., 2021 ). Such overtrust increases the likelihood of misuse of AI by underestimating the potential risks posed by AI (Aroyo et al., 2021 ). This can ultimately lead to excessive trust and reliance on AI (Klingbeil et al., 2024 ). According to Social Cognitive Theory (SCT) (Bandura, 1997 ), human behavior results from the interaction of cognitive characteristics, behaviors, and environmental factors. Based on this theory, self-efficacy is one of the factors that can influence human overtrust in AI. Self-efficacy refers to individuals’ beliefs in their own ability to perform tasks successfully (Bandura, 1997 ). Individuals with low self-efficacy are more likely to exhibit automation bias on machines and AI due to a lack of confidence in their own judgment (Patton, 2023 ; Zhang et al., 2024 ; Lee & Moray, 1994 ). Consequently, lower self-efficacy is associated with greater AI overtrust (Wagner et al., 2018). H1: Students with lower self-efficacy will exhibit greater AI overtrust. Meanwhile, Self-Determination Theory (SDT) posits that an individual's behavior is regulated according to the level of self-determination. Self-determination is composed of basic psychological needs: competence, autonomy, and relatedness (Deci & Ryan, 1985 ). According to Self-Determination Theory, when self-determination is satisfied, extrinsic motivation is initially driven by external incentives (Deci & Ryan, 1985 ), and subsequently develops into intrinsic motivation, which stems from interest in the activity itself (Ryan & Deci, 2000 ; Roca & Gagné, 2008 ). For individuals with strong intrinsic motivation, the intervention of autonomy-based cognitive forcing functions can reduce overtrust in AI (Buçinca et al., 2021 ; Klingbeil et al., 2024 ). In contrast, those primarily driven by extrinsic motivation may be more prone to overtrusting AI due to a lack of critical appraisal of the AI technology (Gregor & Benbasat, 1999 ; Eisbach et al., 2023 ). H2: Students with higher intrinsic motivation will exhibit lower AI overtrust. H3: Students with higher extrinsic motivation will exhibit greater AI overtrust. 2.2 AI literacy as mediator AI literacy refers not only to the technical ability to operate AI but also to the comprehensive capacity to critically understand them from an ethical perspective (Černý, 2024 ). While literacy in traditional education primarily referred to basic skills such as reading, writing, and arithmetic (Černý, 2024 ; Yim & Su, 2024 ), in contemporary society, the concept has expanded to encompass a broad understanding of and ability to navigate digital and intelligent information environments (Yim & Su, 2024 ; Kong, 2014 ). AI literacy comprises three key subcomponents. First, AI critical thinking involves the recognition of the limitations and potential errors of AI and the ability to avoid blindly accepting their outputs (Kong, 2014 ; Dwivedi et al., 2023 ). Second, AI ethics refers to an individual’s capacity to reflect on the social and ethical implications of AI, including concerns about data bias, privacy violations, and source reliability (Ng et al., 2022). Third, AI operation skill denotes the practical and technical proficiency in effectively operating AI tools and systems—an essential competency for addressing digital inequality and improving access to information (Ng et al., 2022; Kong, 2014 ). A lack of AI literacy—specifically, the inability to recognize the limitations and biases of AI and to critically reflect on its potential ethical implications for society—can lead to overtrust in AI. Furthermore, insufficient understanding of or limited ability to use AI may also result in excessive reliance on and overtrust in AI (Kong, 2014 ; Dwivedi et al., 2023 ; Lee & See, 2004 ). In general, high self-efficacy among learners is associated with ethical behavior (Stenmark et al., 2021 ) and plays a critical role in the development of critical thinking skills (Dehghani et al., 2011). In particular, learners with a high level of autonomy—a key driver of intrinsic motivation—tend to adopt a critical perspective on AI through reflective cognition, thereby enhancing their AI literacy (He & Li, 2023 ; Shen & Cui, 2024 ). In contrast, while extrinsic motivation may be effective for achieving short-term goals or task performance, it has limitations in fostering a deeper understanding of AI technologies, ethical reasoning, and epistemological reflection (Artemova, 2024 ). H4: AI literacy mediates the relationship between self-efficacy and AI overtrust. H5: AI literacy mediates the relationship between intrinsic motivation and AI overtrust. H6: AI literacy mediates the relationship between extrinsic motivation and AI overtrust. 2.3 Performance expectation as mediator Performance expectation refers to the degree to which an individual believes that using AI technology will enhance their performance (Venkatesh et al., 2003 ; Dwivedi et al., 2019). It is a key factor that significantly influences the intention to use technology (Venkatesh et al., 2003 ), and a high level of performance expectation may lead to strong trust in AI technology (Harbarth et al., 2025 ). Psychological factors that affect learners’ performance expectations include self-efficacy and motivation. For instance, the more learners believe they can successfully perform academic tasks, the more strongly they believe in their ability to achieve desired learning outcomes (Lent et al., 2008 ). Based on this theoretical background, the present study hypothesizes that the overtrust on AI stemming from learners’ psychological factors is not only mediated by performance expectation but also exerts a sequential mediating effect in conjunction with AI literacy. H7: Performance expectation will mediate the relationship between self-efficacy and AI overtrust. H8: Performance expectation will mediate the relationship between intrinsic motivation and AI overtrust. H9: Performance expectation will mediate the relationship between extrinsic motivation and AI overtrust. H10: Performance expectation and AI literacy will sequentially mediate the relationship between learners’ psychological traits and AI overtrust. H10.1: Performance expectation and AI literacy will sequentially mediate the relationship between learners’ intrinsic motivation and AI overtrust. H10.2: Performance expectation and AI literacy will sequentially mediate the relationship between learners’ extrinsic motivation and AI overtrust. H10.3: Performance expectation and AI literacy will sequentially mediate the relationship between learners’ self-efficacy and AI overtrust. The research model summarizing these hypothesized relationships is presented in Fig. 1 . 3. Materials and methods 3.1 Participants This study collected data through an online survey conducted in Mar 2025 via Embrain, a well-established online survey platform in South Korea. The study received prior approval from the institutional review board of our university. The participants consisted of currently enrolled university students in South Korea, including undergraduate, master’s, and doctoral students. Selection criteria required that all participants had prior experience using ChatGPT for academic-related purposes, such as completing coursework, studying, or conducting research. A total of 300 students participated in the survey. Table 1 presents the demographic characteristics of the participants. Table 1 Demographic Information Frequency Percent Sex Male 150 50.0 Female 150 50.0 Age Group 19 ~ 24 174 58.0 25 ~ 29 97 32.3 30 ~ 34 22 7.3 35~ 7 2.3 Degree Undergraduate Program 238 79.3 Master’s Program 35 11.7 Doctoral Program 27 9.0 Major Humanities 54 18.0 Social Sciences 68 22.7 Engineering 91 30.3 Natural Sciences 43 14.3 Arts and Physical Education 21 7.0 Others 23 7.7 ChatGPT Usage Frequency Daily 57 19.0 4–6 times a week 60 20.0 1–3 times a week 98 32.7 1–3 times a month 69 23.0 Almost never use 16 5.3 3.2 Measurements 3.2.1 Academic self-efficacy (Cronbach’s α = 0.79) To measure students’ academic self-efficacy, the academic self-efficacy questionnaire developed by Nielsen et al. ( 2018 ) was used. This questionnaire includes statements such as: "I believe that with sufficient effort throughout the semester, I can solve academic difficulties by myself," "I believe that with enough study, I can understand complex concepts," and "I believe that even as the assignment deadline approaches, I can manage stress and meet the deadline." This instrument has been widely used in previous studies (Akanni & Oduaran, 2018 ; Hitches et al., 2022 ). Additionally, Naiseh et al. ( 2025 ) explored the impact of academic self-efficacy on AI usage, showing that individuals with higher self-efficacy are more confident in using AI tools effectively for academic tasks. 3.2.2 Motivation (intrinsic motivation Cronbach’s α = 0.75, extrinsic motivation Cronbach’s α = 0.58) Learners’ academic motivation was assessed using items adapted from the Self-Determination Theory (Deci & Ryan, 1985 ), distinguishing between intrinsic (e.g., curiosity-driven learning) and extrinsic (e.g., grade-oriented effort) motivations. Items reflected both autonomy-driven and performance-driven attitudes towards learning, as suggested by prior studies integrating motivational constructs into technology acceptance (Nikou & Economides, 2017 ; Lee et al., 2005 ). 3.2.3 Performance expectations (Cronbach’s α = 0.82) Performance expectations were measured using three items adapted from Aronson and Carlsmith ( 1962 ). These included statements such as “I think using ChatGPT will help me solve problems more easily when studying,” “I believe ChatGPT will provide helpful information for my learning,” and “I think using ChatGPT will save me time when studying.” The reliability of this scale has been previously confirmed (Nan et al., 2022 ; Zhang et al., 2023 a). Additionally, Venkatesh et al. ( 2003 ) explored the impact of performance expectations on attitudes toward technology in the context of UTAUT, highlighting that performance expectations significantly influence users' attitudes toward technology use. 3.2.4 AI literacy (Cronbach’s α = 0.77) AI literacy was measured using items developed from prior research (e.g., Ng et al., 2012; Imjai et al., 2025 ; Lin et al., 2024 ), capturing participants’ understanding of AI fundamentals, practical AI proficiency, and awareness of ethical implications. The items covered cognitive and metacognitive aspects (Kong, 2014 ), including the ability to critically analyze AI-generated information, assess data sources, and adapt AI tools effectively in learning contexts. Participants were also asked to evaluate AI’s limitations, data use transparency, and social accountability, reflecting contemporary AI ethics frameworks. 3.2.5 AI overtrust (Cronbach’s α = 0.85) To assess the degree of overtrust on AI ChatGPT in academic decision-making contexts, an AI overtrust scale was developed based on prior research (e.g., Ullrich et al., 2021 ; Kaur et al., 2022 ; Klingbeil et al., 2024 ). The scale comprised three items, such as “I believe that ChatGPT provides more accurate judgments than human instructors,” measuring users' tendency to overestimate the reliability and usefulness of AI-generated information. This construct reflects miscalibrated trust, where the perceived capabilities of AI exceed its actual performance in learning situations (Lee & See, 2004 ). All items were rated on a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree). 3.3 Data Analysis Strategy Statistical analysis was performed using IBM SPSS Statistics version 27. To test the mediation effects, the PROCESS macro version 4.2 (Hayes, 2012 ) was utilized, specifically using Model 4 to analyze the mediating role of performance expectation and AI literacy in the relationship between learner characteristics such as academic self-efficacy and academic motivation, and AI overtrust. The indirect effects were tested using 5,000 bootstrapping iterations with a 95% bias-corrected confidence interval. If the confidence interval of the indirect effect did not contain 0, the mediation effect was considered statistically significant (Preacher & Hayes, 2008 ). All variables underwent mean-centering to minimize multicollinearity issues. Prior to the analysis, the basic assumptions, including linearity and homoscedasticity, were checked, and all assumptions were met. Additionally, a preliminary review of the relationships between variables and multicollinearity indicated that no multicollinearity was present. 4. Results Table 2 Descriptive statistics and correlation coefficients AOT SE PE AILit AOT 1 - - - SE 0.230** 1 - - PE 0.360** 0.216** 1 - AILit 0.051 0.471** 0.457** 1 Mean 3.16 3.90 4.07 4.01 SD 4.72 0.64 0.65 0.47 Note: AOT, AI overtrust; SE, Self efficacy; PE, Performance Expectation; AILit, AI literacy; Confidence Interval; *p < 0.05, **p < 0.001, *** p = 0.000. Table 3 Regression Coefficients for the Serial Mediation of Self-Efficacy on AI overtrust Dependent variable Independent variable β SE (HC3) t 95% CI PE SE .2261 .0695 3.34** [.0952, .3690] AILit SE .3823 .0415 6.82*** [.2016, .3651] PE .3770 .0509 5.35*** [.1720, .3724] AOT PE .4563 .0794 8.25*** [.4994, .8121] AILit –.2847 .1272 –4.45*** [–.8170, –.3163] Note: PE, Performance Expectation; SE, Self efficacy; AILit, AI literacy; AOT, AI overtrust; β, Standardized regression coefficient; SE(HC3), Heteroscedasticity-consistent standard error (HC3); t, t-statistic; 95% CI, 95% Confidence Interval; *p < 0.05, **p < 0.001, *** p = 0.000. Table 4 Indirect and Contrast Effects of Self-Efficacy on AI overtrust Indirect Path Effect Boot SE 95% CI Ind1: SE → PE → AOT .1522 .0481 [.0648, .2536] Ind2: SE → AILit → AOT –.1606 .0423 [–.2492, –.0838] Ind3: SE → PE → AILit → AOT –.0358 .0145 [–.0691, –.0131] Indirect Effect Contrasts C1: Ind1 – Ind2 .3128 .0700 [.1843, .4605] C2: Ind1 – Ind3 .1880 .0591 [.0804, .3123] C3: Ind2 – Ind3 –.1248 .0379 [–.2039, –.0567] Note: PE, Performance Expectation; SE, Self efficacy; AILit, AI literacy; AOT, AI overtrust; β, Standardized regression coefficient; Effect = Indirect effect estimate; Boot SE = Bootstrapped standard error of the indirect effect; 95% CI, 95% Confidence Interval Table 5 Summary of Effects of Self-Efficacy on AI overtrust Path (SE → AOT) Effect 95% CI Total Effect 0.2971 - Direct Effect 0.3413 [.1674, .5152] Indirect Effect –.0442 [–.1494, .0696] Table 6 Regression Coefficients for the Serial Mediation of Motivation on AI overtrust Panel A. Intrinsic Motivation (IM) Dependent variable Independent variable β SE (HC3) t 95% CI PE IM .1304 .0539 2.42* [.0242, .2365] AILit IM .0759 .0348 2.18* [.0073, .1444] PE .3194 .0552 5.79*** [.2108, .4281] AOT PerEx .6407 .0810 7.91*** [.4813, .8000] AILit − .4033 .1312 -3.07** [-.6616, − .1451] Panel B. Extrinsic Motivation (EM) Dependent variable Independent variable β SE (HC3) t 95% CI PE EM .3147 .0605 5.21*** [.1957, .4337] AILit EM .0298 .0495 0.60 [-.0676, .1271] PE .3239 .0619 5.23*** [.2021, .4457] AOT PerEx .5715 .0823 6.94*** [.4095, .7334] AILit − .3742 .1277 -2.93** [-.6256, − .1229] Note: PE, Performance Expectation; IM, Intrinsic Motivation; EM, Extrinsic Motivation; AILit, AI literacy; AOT, AI overtrust; β, Standardized regression coefficient; SE(HC3), Heteroscedasticity-consistent standard error (HC3); t, t-statistic; 95% CI, 95% Confidence Interval; *p < 0.05, **p < 0.001, *** p = 0.000. Table 7 Indirect and Contrast Effects of Motivation on AI overtrust Panel A. Intrinsic Motivation (IM) Indirect Path Effect Boot SE 95% CI Ind1: IM → PE → AOT .0835 .0352 [.0178, .1538] Ind2: IM → AILit → AOT − .0306 .0173 [-.0701, − .0031] Ind3: IM → PE → AILit → AOT − .0168 .0096 [-.0393, − .0024] Indirect Effect Contrasts C1: Ind1 – Ind2 .1141 .0389 [.0444, .1957] C2: Ind1 – Ind3 .1003 .0429 [.0215, .1882] C3: Ind2 – Ind3 − .0138 .0183 [-.0533, .0194] Panel B. Extrinsic Motivation (EM) Indirect Path Effect Boot SE 95% CI Ind1: EM → PE → AOT .1799 .0408 [.1026, .2621] Ind2: EM → AILit → AOT − .0111 .0193 [-.0527, .0262] Ind3: EM → PE → AILit → AOT − .0382 .0161 [-.0748, − .0130] Indirect Effect Contrasts C1: Ind1 – Ind2 .1910 .0417 [.1111, .2735] C2: Ind1 – Ind3 .2180 .0506 [.1236, .3220] C3: Ind2 – Ind3 .0270 .0275 [-.0184, .0904] Note: PE, Performance Expectation; IM, Intrinsic Motivation; EM, Extrinsic Motivation; AILit, AI literacy; AOT, AI overtrust; β, Standardized regression coefficient; Effect = Indirect effect estimate; Boot SE = Bootstrapped standard error of the indirect effect; 95% CI, 95% Confidence Interval Table 8 Summary of Effects of Motivation on AI overtrust Effect Type Intrinsic Motivation 95% CI (Intrinsic) Extrinsic Motivation 95% CI (Extrinsic) Total Effect 0.2184 - 0.4094 - Direct Effect 0.1823 [0.0544, 0.3101] 0.2788 [0.1304, 0.4272] Indirect Effect 0.0361 [-0.0342, 0.1024] 0.1306 [0.0459, 0.2138] 4.1 Descriptive Statistics and Correlations Descriptive statistics and correlation coefficients for the study variables are presented in Table 2 . AI overtrust (AOT) was positively correlated with self-efficacy (SE) (r = .230, p < .001) and performance expectation (PE) (r = .360, p < .001), while AI literacy (AILit) was not significantly correlated with AOT. Table 3 displays the regression coefficients for the serial mediation model of SE on AOT. SE had a significant positive effect on PE (β = .2261, p < .001) and AILit (β = .3823, p = .000), and PE also had a significant effect on AILit (β = .3770, p = .000). In turn, AILit had a significant negative effect on AOT (β = –.2847, p = .000), while PE had a significant positive direct effect on AOT (β = .4563, p = .000). These results suggest that higher SE increases PE and AILit, with AILit reducing AOT and PE directly contributing to it. 4.2 Direct Effects of Psychological Predictors on AI overtrust The analysis of the direct effects of SE and motivation variables (intrinsic motivation, extrinsic motivation) on AOT revealed that learners’ SE had a significant positive impact on AOT (β = .3413, p < .01) (Table 5 ), which contradicts hypothesis H1. This result also diverges from prior research (Patton, 2023 ; Zhang et al., 2024 ; Lee & Moray, 1994 ), which suggested that lower SE is associated with higher levels of AOT. Intrinsic motivation (IM) had a significant positive effect on AOT (β = .1823, p < .01) (Table 8 ), contradicting hypothesis H2. Hypothesis H2 predicted that students with higher IM would exhibit lower levels of AOT; however, the actual findings revealed the opposite. Similarly, extrinsic motivation (EM) also showed a significant positive effect on AOT (β = .2788, p < .001) (Table 8 ), supporting hypothesis H3. This indicates that students who are motivated by external rewards tend to have higher expectations regarding the functionality and results of AI, which in turn leads to increased AOT. The structural model for these relationships is shown in Fig. 3 , where both IM and EM have significant positive effects on PE, which in turn influences AOT directly and indirectly through AILit. 4.3 Mediating Role of Performance Expectation and AI Literacy The structural model of SE’s effects on AOT via PE and AILit is shown in Fig. 2 . As illustrated, SE positively affected both PE and AILit, PE positively affected AOT, and AILit negatively affected AOT. Next, we examined the direct and sequential mediating effects of PE and AILit. The indirect path from SE through PE to AOT showed a significant positive mediating effect (β = .1522, Boot CI = [.0648, .2536]) (Table 4 ), supporting H7. In contrast, the path from SE → AILit → AOT showed a significant negative mediating effect (β = –.1606, CI = [–.2492, –.0838]) (Table 4 ), supporting H4 and suggesting that AILit can suppress AOT. IM also had a positive indirect effect on AOT through PE (β = .0835, CI = [.0178, .1538]) (Table 7 ), supporting H8, while the path via AILit showed a significant negative effect (β = –.0306, CI = [–.0701, –.0031]) (Table 7 ), supporting H5. For EM, the path via PE showed a significant positive indirect effect (β = .1799, CI = [.1026, .2621]) (Table 7 ), supporting H9. The detailed regression coefficients for these serial mediation models of IM and EM are presented in Table 6 . The complex path from PE to AILit and then to critical thinking showed a significant negative indirect effect (β = –.0382, CI = [–.0748, –.0130]) (Table 7 ), supporting H6. However, the single path via only AILit (β = –.0111) was not statistically significant, indicating that the direct path of H6 was not supported. Overall, these results suggest that SE and IM form pathways through AILit that critically mediate AOT, while EM does not show a significant mediating effect when AILit is considered alone. 4.4 Serial Mediation Effects: Performance expectation and AI Literacy The analysis of the sequential mediating effects of PE and AILit revealed significant indirect paths for all psychological independent variables. For SE, a significant negative indirect effect was found in the path from SE through PE and AILit to AOT (β = –.0358, CI = [–.0691, –.0131]) (Table 4 ). This suggests that high SE increases PE, which leads to improved AILit, ultimately enhancing critical judgment toward AI and suppressing overtrust (supporting H10.3). A similar trend was observed for IM (β = –.0168, CI = [–.0393, –.0024]) (Table 7 ), supporting H10.1. For EM, the indirect effect in the sequential mediating path was the strongest (β = –.0382, CI = [–.0748, –.0130]) (Table 7 ), supporting H10.2. This result indicates that learners with EM, who prioritize external rewards, can also suppress overtrust in AI technologies if they sufficiently develop AILit. 5. Discussion This study analyzed the impact of psychological factors such as self-efficacy, intrinsic motivation, and extrinsic motivation on AI overtrust, and examined the mediating and sequential mediating roles of performance expectation and AI literacy in this pathway. The key findings derived from this study are summarized in four main points. First, higher self-efficacy was found to indirectly increase the level of AI overtrust. This result contradicts previous studies that argued lower self-efficacy leads to greater AI overtrust (Patton, 2023 ; Zhang et al., 2024 ; Lee & Moray, 1994 ; Lee & See, 2004 ). Lee and See ( 2004 ) suggested that users with low confidence in their ability to use automated technologies tend to overtrust them. Similarly, Zhang et al. ( 2024 ) found that learners with low self-efficacy are more likely to become dependent on AI. In contrast, the current analysis reveals an indirect pathway whereby high self-efficacy in learning leads to increased performance expectation, which in turn contributes to greater AI overtrust. This suggests that learners with strong self-efficacy may develop overly optimistic expectations about AI’s ability to enhance learning outcomes, leading to an inflated perception of AI capabilities. Such overestimation may foster excessive trust in AI (Lent et al., 2008 ). Second, higher levels of intrinsic motivation were found to directly increase AI overtrust. This finding contrasts with prior studies that argued intrinsic motivation enhances cognitive engagement and critical thinking, thereby reducing overreliance on AI (Buçinca et al., 2021 ; Klingbeil et al., 2024 ). The present result suggests that intrinsically motivated learners, while typically engaging in deeper cognitive processing, may develop emotional dependence on AI tools such as ChatGPT due to their repetitive and rewarding interaction patterns. This emotional attachment—ironically rooted in high cognitive engagement—can undermine critical thinking and foster uncritical acceptance of AI outputs (Yankouskaya et al., 2025 ). Third, extrinsic motivation had a stronger direct and indirect effect on AI overtrust than intrinsic motivation. This may be attributed to the instrumental and performance-oriented learning context fostered by AI tools such as ChatGPT, which amplifies the influence of extrinsic motivation (Klingbeil et al., 2024 ). Learners with high extrinsic motivation tend to view AI as a means to maximize learning efficiency, which elevates their performance expectation. This, in turn, leads to uncritical acceptance of AI outputs and fosters AI overtrust. Fourth, performance expectation and AI literacy function as key mediators in the formation of AI overtrust based on learners’ psychological traits. The results show that when learners expect high academic efficiency from AI use, they tend to accept AI outputs uncritically. In contrast, AI literacy plays a crucial role in mitigating or blocking this path by enabling learners to understand the limitations of AI—such as algorithmic bias, data opacity, and lack of accountability—and to critically evaluate AI-generated information (Ng et al., 2021 ). Notably, while extrinsic motivation alone did not significantly reduce AI overtrust through AI literacy, it exhibited a strong negative effect when mediated sequentially by performance expectation and AI literacy. This suggests that extrinsically motivated learners may curb AI overtrust if their performance expectations lead to the development of AI literacy. These findings suggest that extrinsic motivation does not inherently lead to AI overtrust, but that cognitive mediation processes involving motivation, performance expectation, and AI literacy play a critical role in shaping this relationship. Overall, this study highlights the risk that learners’ psychological characteristics can lead to overtrust in AI and proposes new directions for AI literacy education and motivation-based learning strategies in AI-integrated environments. 6. Implications and Limitations The results offer several important implications for educational practice and policy. First, from a theoretical perspective, this study empirically demonstrates that AI literacy and performance expectation can function as a complex mediating structure explaining AI overtrust. While prior studies have focused on the impact of psychological factors on attitudes toward AI, this study offers new insights by clarifying how these factors lead to AI overtrust, emphasizing AI literacy as a key cognitive mediator in the process. From a practical standpoint, as the educational use of generative AI tools, including ChatGPT, continues to grow, this study suggests that the mere introduction of such technologies may result in unintended side effects, such as AI overtrust. Learners with high self-efficacy and strong intrinsic and extrinsic motivation are more likely to trust and use AI. However, when their AI literacy is low, this trust may easily develop into overtrust. Therefore, it is essential to provide adequate pre-usage education tailored to learners’ psychological characteristics. Such education should foster proper AI literacy while also enhancing awareness of AI’s limitations and potential errors, as well as learners’ ability to critically assess AI outputs. Learners driven by extrinsic motivations, such as grades or financial rewards, are particularly vulnerable to overtrust. This highlights the need for a structured AI literacy curriculum targeting these groups. This study also presents several limitations. First, the sample was limited to university students in South Korea, which restricts the generalizability of the findings. Future research should include more diverse participant groups across age, cultural, and occupational backgrounds. Second, the cross-sectional design limits the ability to establish causal relationships among variables. Longitudinal studies are needed to explore how psychological factors and overtrust evolve with continued AI use. Third, all variables were measured through self-report questionnaires, which may be influenced by social desirability bias or subjective interpretation. Future studies should consider incorporating behavioral data or AI usage logs to improve data reliability. In conclusion, this study identifies the potential for learners’ psychological characteristics to lead to AI overtrust in educational settings. It also empirically demonstrates that AI literacy plays a critical mediating role in this process. These findings offer foundational insights for future AI integration in education and for the development of related policy frameworks. Specifically, the study emphasizes that AI literacy education must extend beyond technical skills. It should incorporate critical thinking, an understanding of algorithmic transparency, and awareness of data ethics and social responsibility. These elements are essential and should be meaningfully integrated into educational practice. 7. Future Research Directions This study is significant in empirically identifying the psychological factors influencing AI overtrust and the mediating factors involved. However, to better understand these results in a more refined and expanded form, follow-up research from various perspectives is necessary. First, since this study was conducted as a cross-sectional study using data from a specific point in time, it is difficult to clearly identify changes over time or causal flows between the variables. Future research should use a longitudinal design to track how learners’ AI usage experiences affect their self-efficacy and motivation, and how these changes, in turn, influence the level of AI overtrust over the long term. Additionally, the sample in this study was limited to university students in South Korea, meaning that cultural context, age, education level, and personality traits were not considered. Therefore, future studies should expand the research to include learner groups from diverse cultural backgrounds and analyze whether the impact of psychological factors on AI overtrust varies across cultures. Particularly in countries with different social perceptions of AI, technological accessibility, or educational environments, the results from this study may differ. Furthermore, this study was based on self-reported surveys, which have the limitation of relying on subjective perceptions. In the future, it would be beneficial to collect a combination of quantitative and qualitative data, such as actual AI tool usage logs, learning performance indicators, and objective literacy assessment tools, to derive more accurate and comprehensive results. For example, analyzing the quality of texts written by learners using ChatGPT or examining judgment errors in the way tasks are solved could empirically confirm the impact of AI overtrust on the actual learning process. Lastly, this study analyzed the mediating effects focusing on performance expectation and AI literacy. However, future studies should include a broader range of cognitive and emotional factors, such as AI trust, the ability to assess the authenticity of information, and tolerance for ambiguity, to construct a more multi-layered explanatory model of AI overtrust. Moreover, research that analyzes the perceptions and usage of AI by not only learners but also teachers, educators, and policymakers is required to enhance the ecological understanding of educational AI usage. In conclusion, as AI technology rapidly spreads, research that goes beyond analyzing learner-centered psychological factors and encompasses various contexts, groups, and long-term effects will significantly contribute to the practical strategies for improving the quality of AI education and preventing overtrust. Abbreviations AI: Artificial Intelligence LLM: Large Language Model AOT: AI Overtrust SE: Self-Efficacy IM: Intrinsic Motivation EM: Extrinsic Motivation PE: Performance Expectation AILit: AI Literacy CI: Confidence Interval SE (HC3): Standard Error (Heteroscedasticity-consistent type 3) Boot SE: Bootstrapped Standard Error Declarations Acknowledgements Not applicable. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Author information Authors and Affiliations Department of Interaction Science, Department of Human-Artificial Intelligence Interaction, Sungkyunkwan University, 03063 Seoul, Republic of Korea Youngsang Kim & Jang Hyun Kim Institute of Chinese Language and Culture Education, College of Chinese Language and Culture, Huaqiao University, Xiamen, China Shunan Zhang Contributions Youngsang Kim: conceptualization, data analysis, and original draft writing; Shunan Zhang: conceptualization and supervision; Jang Hyun Kim: conceptualization, supervision, funding acquisition, and manuscript review and editing. Corresponding author: Correspondence to Shunan Zhang Ethics declarations Ethics approval and consent to participate All procedures involving human participants in this study were conducted in accordance with the ethical standards of the Declaration of Helsinki. The study protocol was reviewed and approved by the Institutional Review Board of Sungkyunkwan University (IRB approval number: 2025-02-081). Written informed consent was obtained from all participants prior to their inclusion in the study. Consent for publication Written informed consent for publication of anonymized data was obtained from all participants prior to their inclusion in the study. Competing interests The author declares no conflict of interest Data Availability The data and material used to derive the findings in this study are available from the corresponding author upon reasonable request. References Acosta-Enriquez, B. G., Ballesteros, M. A. A., Valle, M. D. L. A. G., Angaspilco, J. E. M., Lalupú, J. D. R. A., Jaico, J. L. B., ... & Castillo, W. E. J. (2025). The mediating role of academic stress, critical thinking and performance expectations in the influence of academic self-efficacy on AI dependence: case study in college students. Computers and Education: Artificial Intelligence 8 , 100381. https://doi.org/10.1016/j.caeai.2025.100381 Agbong-Coates, I. J. G. (2024). ChatGPT integration significantly boosts personalized learning outcomes: A Philippine study. International journal of educational management and development studies, 5(2) , 165-186. https://doi.org/10.53378/353067 Akanni, A. A., & Oduaran, C. A. (2018). Perceived social support and life satisfaction among freshmen: Mediating roles of academic self-efficacy and academic adjustment. Journal of Psychology in Africa, 28(2) , 89-93. https://doi.org/10.1080/14330237.2018.1454582 Alzain, E., Nagi, K. A., & Algobaei, F. (2024). The Quality of Google Translate and ChatGPT English to Arabic Translation: The Case of Scientific Text Translation. In Forum for Linguistic Studies , 6(3) , pp. 837-849). https://doi.org/10.30564/fls.v6i3.6799 Aronson, E., & Carlsmith, J. M. (1962). Performance expectation as a determinant of actual performance . The Journal of Abnormal and Social Psychology, 65(3) , 178. https://doi.org/10.1037/h0042291 Aroyo, A. M., De Bruyne, J., Dheu, O., Fosch-Villaronga, E., Gudkov, A., Hoch, H., ... & Tamò-Larrieux, A. (2021). Overtrusting robots: Setting a research agenda to mitigate overtrust in automation. Paladyn, Journal of Behavioral Robotics, 12(1) , 423-436. http://doi.org/10.1515/pjbr-2021-0029 Artemova, I. (2024). Bridging Motivation and AI in Education: An Activity Theory Perspective. Digital Education Review, 45 , 59-69. https://doi.org/ 10.1344/der.2024.45.59-67 Bandura, A., & Wessels, S. (1997). Self-efficacy (pp. 4-6) . Cambridge: Cambridge University Press. Bettayeb, A. M., Abu Talib, M., Sobhe Altayasinah, A. Z., & Dakalbab, F. (2024). Exploring the impact of ChatGPT: conversational AI in education. Frontiers in Education, 9 , 1379796. https://doi.org/10.3389/feduc.2024.1379796. Bhullar, P. S., Joshi, M., & Chugh, R. (2024). ChatGPT in higher education-a synthesis of the literature and a future research agenda. Education and Information Technologies 29 , 1-22. http://doi.org/10.1007/s10639-024-12723-x. Buçinca, Z., Malaya, M. B., & Gajos, K. Z. (2021). To trust or to think: cognitive forcing functions can reduce overreliance on AI in AI-assisted decision-making. Proceedings of the ACM on Human-computer Interaction,5(CSCW1) , 1-21. http://doi.org/10.1145/3449287. Černý, M. (2024). University students’ conceptualisation of AI literacy: Theory and empirical evidence. Social Sciences, 13(3) , 129. http://doi.org/10.3390/socsci13030129. Deci, E. L., & Ryan, R. M. (1985). Conceptualizations of intrinsic motivation and self-determination . Intrinsic motivation and self-determination in human behavior , 11-40. http://doi.org/10.1007/978-1-4899-2271-7_2. Dwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., ... & Wright, R. (2023). Opinion Paper: “So what if ChatGPT wrote it?” Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy. International Journal of Information Management , 71, 102642. http://doi.org/10.1016/j.ijinfomgt.2023.102642. Eisbach, S., Langer, M., & Hertel, G. (2023). Optimizing human-AI collaboration: Effects of motivation and accuracy information in AI-supported decision-making. Computers in Human Behavior: Artificial Humans,1(2), 100015. http://doi.org/10.1016/j.chbah.2023.100015. Gregor, S., & Benbasat, I. (1999). Explanations from intelligent systems: Theoretical foundations and implications for practice. MIS quarterly , 23(4) , 497-530. http://doi.org/10.2307/249487. Harbarth, L., Gößwein, E., Bodemer, D., & Schnaubert, L. (2025). (Over) trusting AI recommendations: How system and person variables affect dimensions of complacency. International Journal of Human–Computer Interaction, 41(1) , 391-410. http://doi.org/10.1080/10447318.2023.2301250. Hasanein, A. M., & Sobaih, A. E. E. (2023). Drivers and consequences of ChatGPT use in higher education: Key stakeholder perspectives. European journal of investigation in health, psychology and education, 13(11), 2599-2614. http://doi.org/10.3390/ejihpe13110181. Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable mediation, moderation, and conditional process modeling. He, L., & Li, C. (2023). Continuance intention to use mobile learning for second language acquisition based on the technology acceptance model and self-determination theory. Frontiers in Psychology, 14 , 1185851. https://doi.org/10.3389/fpsyg.2023.1185851. Hitches, E., Woodcock, S., & Ehrich, J. (2022). Building self-efficacy without letting stress knock it down: Stress and academic self-efficacy of university students . International Journal of Educational Research Open, 3 , 100124. http://doi.org/10.1016/j.ijedro.2022.100124. Imjai, N., Yordudom, T., Yaacob, Z., Saad, N. H. M., & Aujirapongpan, S. (2025). Impact of AI literacy and adaptability on financial analyst skills among prospective Thai accountants: The role of critical thinking. Technological Forecasting and Social Change, 210 , 123889. https://doi.org/10.1016/j.techfore.2024.123889. Jo, H. (2024). From concerns to benefits: a comprehensive study of ChatGPT usage in education. International Journal of Educational Technology in Higher Education, 21(1) , 35. http://doi.org/10.1186/s41239-024-00471-4. Kaur, D., Uslu, S., Rittichier, K. J., & Durresi, A. (2022). Trustworthy artificial intelligence: a review. ACM computing surveys (CSUR), 55(2) , 1-38. https://doi.org/10.1145/3491209. Klingbeil, A., Grützner, C., & Schreck, P. (2024). Trust and reliance on AI—An experimental study on the extent and costs of overreliance on AI. Computers in Human Behavior, 160 , 108352. http://doi.org/10.1016/j.chb.2024.108352, Kong, S. C. (2014). Developing information literacy and critical thinking skills through domain knowledge learning in digital classrooms: An experience of practicing flipped classroom strategy. Computers & education, 78 , 160-173. http://doi.org/10.1016/j.compedu.2014.05.009. Kovari, A. (2025). Ethical use of ChatGPT in education—Best practices to combat AI-induced plagiarism. In Frontiers in Education. Frontiers Media SA , 9 , 1465704, https://doi.org/10.3389/feduc.2024.1465703. Laitinen, S., Lepola, J., & Vauras, M. (2017). Early motivational orientation profiles and language comprehension skills: From preschool to Grade 3. Learning and Individual Differences, 53 , 69-78. Lee, J. D., & Moray, N. (1994). Trust, self-confidence, and operators’ adaptation to automation. International journal of human-computer studies, 40(1) , 153-184. https://doi.org/10.1006/ijhc.1994.1007. Lee, J. D., & See, K. A. (2004). Trust in automation: Designing for appropriate reliance. Human factors, 46(1), 50-80. https://doi.org/10.1518/hfes.46.1.50_30392. Lee, M. K., Cheung, C. M., & Chen, Z. (2005). Acceptance of Internet-based learning medium: the role of extrinsic and intrinsic motivation. Information & management, 42(8), 1095-1104. Lent, R. W., Sheu, H. B., Singley, D., Schmidt, J. A., Schmidt, L. C., & Gloster, C. S. (2008). Longitudinal relations of self-efficacy to outcome expectations, interests, and major choice goals in engineering students. Journal of vocational behavior, 73(2) , 328-335. https://doi.org/10.1016/j.jvb.2008.07.005. Li, Z., Wang, C., & Bonk, C. J. (2024). Exploring the Utility of ChatGPT for Self-Directed Online Language Learning. Online Learning, 28(3), 157-180. https://eric.ed.gov/?id=EJ1446866. Lin, X. F., Zhou, Y., Shen, W., Luo, G., Xian, X., & Pang, B. (2024). Modeling the structural relationships among Chinese secondary school students’ computational thinking efficacy in learning AI, AI literacy, and approaches to learning AI. Education and Information Technologies, 29(5) , 6189-6215. http://doi.org/10.1007/s10639-023-12029-4. Mai, D. T. T., Da, C. V., & Hanh, N. V. (2024). The use of ChatGPT in teaching and learning: a systematic review through SWOT analysis approach. In Frontiers in Education. Frontiers Media SA , 9 , 1327769, http://doi.org/10.3389/feduc.2024.1328769. Mohebi, L. (2024). Empowering learners with ChatGPT: insights from a systematic literature exploration. Discover Education, 3(1) , 36. http://doi.org/10.1007/s44217-024-00120-y. Naiseh, M., Babiker, A., Al-Shakhsi, S., Cemiloglu, D., Al-Thani, D., Montag, C., & Ali, R. (2025). Attitudes Towards AI: The Interplay of Self-Efficacy, Well-Being, and Competency . Journal of Technology in Behavioral Science, 1-14. http://doi.org/10.1007/s41347-025-00486-2. Nan, D., Lee, H., Kim, Y., & Kim, J. H. (2022). My video game console is so cool! A coolness theory-based model for intention to use video game consoles. Technological Forecasting and Social Change, 176 , 121451. http://doi.org/10.1016/j.techfore.2021.121451. Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021). Conceptualizing AI literacy: An exploratory review. Computers and Education: Artificial Intelligence, 2 , 100041. https://doi.org/10.1016/j.caeai.2021.100041. Nielsen, T., Dammeyer, J., Vang, M. L., & Makransky, G. (2018). Gender fairness in self-efficacy? A Rasch-based validity study of the General Academic Self-efficacy scale (GASE). Scandinavian Journal of Educational Research, 62(5) , 664-681. http://doi.org/10.1080/00313831.2017.1306796. Nikou, S. A., & Economides, A. A. (2017). Mobile-Based Assessment: Integrating acceptance and motivational factors into a combined model of Self-Determination Theory and Technology Acceptance. Computers in Human Behavior, 68 , 83-95. https://doi.org/10.1016/j.chb.2016.11.020Get rights and content. Patton, C. E. (2023). The Influence of Trust, Self-Confidence and Task Difficulty on Automation Use (Doctoral dissertation, Colorado State University). Pithers, R. T., & Soden, R. (2000). Critical thinking in education: A review. Educational research, 42(3) , 237-249. https://doi.org/10.1080/001318800440579 Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior research methods, 40(3) , 879-891. http://doi.org/10.3758/BRM.40.3.879. Roca, J. C., & Gagné, M. (2008). Understanding e-learning continuance intention in the workplace: A self-determination theory perspective. Computers in human behavior, 24(4) , 1585-1604. https://doi.org/10.1016/j.chb.2007.06.001. Ryan, R. M., & Deci, E. L. (2000). Intrinsic and extrinsic motivations: Classic definitions and new directions. Contemporary educational psychology, 25(1) , 54-67. https://doi.org/10.1006/ceps.1999.1020. Shen, Y., & Cui, W. (2024). Perceived support and AI literacy: the mediating role of psychological needs satisfaction. Frontiers in Psychology, 15 , 1415248. https://doi.org/10.3389/fpsyg.2024.1415248. Sok, S., & Heng, K. (2023). ChatGPT for education and research: A review of benefits and risks. Cambodian Journal of Educational Research, 3(1) , 110-121. http://dx.doi.org/10.2139/ssrn.4378735 Stenmark, C. K., Redfearn, R. A., & Kreitler, C. M. (2021). Self-efficacy and ethical decision-making. Ethics & Behavior, 31(5) , 301-320. https://doi.org/10.1080/10508422.2020.1776617. Ullrich, D., Butz, A., & Diefenbach, S. (2021). The development of overtrust: An empirical simulation and psychological analysis in the context of human–robot interaction. Frontiers in Robotics and AI, 8 , 554578. http://doi.org/10.3389/frobt.2021.554578. Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS quarterly, 27(3) , 425-478. http://doi.org/10.2307/30036540. Yankouskaya, A., Liebherr, M., & Ali, R. (2025). Can ChatGPT Be Addictive? A Call to Examine the Shift from Support to Dependence in AI Conversational Large Language Models. Human-Centric Intelligent Systems, 5 , 77-89, https://doi.org/10.1007/s44230-025-00090-w Ye, J. H., Zhang, M., Nong, W., Wang, L., & Yang, X. (2025). The relationship between inert thinking and ChatGPT dependence: An I-PACE model perspective. Education and information technologies, 30(3) , 3885-3909. https://doi.org/10.1007/s10639-024-12966-8 Yim, I. H. Y., & Su, J. (2024). Artificial intelligence (AI) learning tools in K-12 education: A scoping review. Journal of Computers in Education, 12(1), 93-131. https://doi.org/10.1007/s40692-023-00304-9 Zhang, S., Che, S., Nan, D., & Kim, J. H. (2023). How does online social interaction promote students’ continuous learning intentions? Frontiers in Psychology , 14, 1098110. http://doi.org/10.3389/fpsyg.2023.1098110. Zhang, S., Zhao, X., Zhou, T., & Kim, J. H. (2024). Do you have AI dependency? The roles of academic self-efficacy, academic stress, and performance expectations on problematic AI usage behavior. International Journal of Educational Technology in Higher Education, 21(1) . https://doi.org/10.1186/s41239-024-00467-0 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 19 Dec, 2025 Reviewers agreed at journal 24 Sep, 2025 Reviews received at journal 21 Sep, 2025 Reviewers agreed at journal 15 Sep, 2025 Reviews received at journal 15 Sep, 2025 Reviewers agreed at journal 15 Sep, 2025 Reviewers invited by journal 15 Sep, 2025 Editor invited by journal 19 Aug, 2025 Editor assigned by journal 12 Aug, 2025 Submission checks completed at journal 12 Aug, 2025 First submitted to journal 07 Aug, 2025 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-7315296","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":518248256,"identity":"2d884ddc-d0e3-4292-9903-73fb672db1f6","order_by":0,"name":"Youngsang Kim","email":"","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Youngsang","middleName":"","lastName":"Kim","suffix":""},{"id":518248257,"identity":"627a9c05-1a26-45f1-9b94-1c24a7180468","order_by":1,"name":"Jang Hyun Kim","email":"","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Jang","middleName":"Hyun","lastName":"Kim","suffix":""},{"id":518248258,"identity":"7e03ef26-d188-4261-a8ea-e8bbe8a3826f","order_by":2,"name":"Shunan Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAElEQVRIiWNgGAWjYDACCQiVACIMEiqAJDNzA/FaCj6cAWlhJEHLx5ltIIqAFvnZPWYPfu6oy+Of3X5xM++82mj+dqCWHxXbcGoxuHPG3LD3DFuxxJ0zxca8247nzjjM2MDYc+Y2bi0SOWYSvG08iQ03ctKAWo7lNgC1MDO24dYiPyPHTPJvm0Ti/Bs56b955xzLnU9IC8ONHDNp3jaDxA030g8Yzmyoyd1ASIvBjbQyadm2hMSNN3IYDD4cO5C7EajlID6/yM9I3ib5tq0ucd6N9AcGCTV1ufPOHz744EcFHochAI8BkDgMZh4gRj0QsD8AEnVEKh4Fo2AUjIKRBAAG4GDxZdlflQAAAABJRU5ErkJggg==","orcid":"","institution":"Huaqiao University","correspondingAuthor":true,"prefix":"","firstName":"Shunan","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2025-08-07 06:23:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7315296/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7315296/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":92009550,"identity":"cac7a21a-24b8-4494-9505-8d63effbe26d","added_by":"auto","created_at":"2025-09-23 15:37:54","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":197316,"visible":true,"origin":"","legend":"","description":"","filename":"ManuscriptBMCfullversion250812.docx","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/74b070a7cd2007a38e1b8ec2.docx"},{"id":92009549,"identity":"750d6d42-4901-4f62-97be-241405e278ee","added_by":"auto","created_at":"2025-09-23 15:37:54","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5813,"visible":true,"origin":"","legend":"","description":"","filename":"026eb486731c491082d3ff10d174734a.json","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/592ad06cf5f997f8426df685.json"},{"id":92009552,"identity":"417ca366-642b-41c7-a9ea-45f5af930aae","added_by":"auto","created_at":"2025-09-23 15:37:54","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":162755,"visible":true,"origin":"","legend":"","description":"","filename":"026eb486731c491082d3ff10d174734a1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/62aea7a1aed4e210aac564a4.xml"},{"id":92008446,"identity":"db34c3ac-858b-4236-a476-76579351cfff","added_by":"auto","created_at":"2025-09-23 15:29:54","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53542,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/a89374ee9217291952f7ee2e.png"},{"id":92008440,"identity":"84301740-6ea2-4340-98d5-110c5ca3599b","added_by":"auto","created_at":"2025-09-23 15:29:54","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24264,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/cc1f255cca50f626881be4e2.png"},{"id":92008448,"identity":"6a17e0b1-27fd-4d1f-93ff-b2461e16e31c","added_by":"auto","created_at":"2025-09-23 15:29:54","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42605,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/a5dd605fd7a87b8b3f7b2172.png"},{"id":92009551,"identity":"d32b5acb-4399-48c2-ba26-a03b78595070","added_by":"auto","created_at":"2025-09-23 15:37:54","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17559,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/c249622d02ada57276a7bd35.png"},{"id":92008443,"identity":"1eb45352-0e3d-424c-98f6-fc509ad0e2e7","added_by":"auto","created_at":"2025-09-23 15:29:54","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10256,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/cc423ec435b4b451141b08d4.png"},{"id":92008444,"identity":"19b322d7-bc6c-4b9a-81bf-92106a007cec","added_by":"auto","created_at":"2025-09-23 15:29:54","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12931,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/1435d3de1244ae9b4f16379e.png"},{"id":92008450,"identity":"cea3cbe8-f848-4f1e-88fe-48d46da73ab2","added_by":"auto","created_at":"2025-09-23 15:29:55","extension":"xml","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":160502,"visible":true,"origin":"","legend":"","description":"","filename":"026eb486731c491082d3ff10d174734a1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/b2896727abaee8821e14bf67.xml"},{"id":92008449,"identity":"718c977c-ae88-47db-84dc-2dba79dbd6b8","added_by":"auto","created_at":"2025-09-23 15:29:55","extension":"html","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":168720,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/77a7262c80c17802996d23b9.html"},{"id":92008437,"identity":"2312e4e7-8464-4d90-8d3e-07bbe0c91c45","added_by":"auto","created_at":"2025-09-23 15:29:54","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":53542,"visible":true,"origin":"","legend":"\u003cp\u003eResearch model\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/77b3356db136e790947206a1.png"},{"id":92008438,"identity":"4b3edc49-6d24-47b8-91b0-815cb4aa144b","added_by":"auto","created_at":"2025-09-23 15:29:54","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":24264,"visible":true,"origin":"","legend":"\u003cp\u003eEffects of self-efficacy on AI overtrust.\u003c/p\u003e\n\u003cp\u003eNote: *p\u0026lt;0.05, **p\u0026lt;0.001, *** p=0.000.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/51ca40c03255da4b6ad97a04.png"},{"id":92008442,"identity":"e80df847-10c1-4576-a906-e9cce39c506a","added_by":"auto","created_at":"2025-09-23 15:29:54","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":42605,"visible":true,"origin":"","legend":"\u003cp\u003eEffects of motivation on AI overtrust.\u003c/p\u003e\n\u003cp\u003eNote: *p\u0026lt;0.05, **p\u0026lt;0.001, *** p=0.000.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/f855813227a8c5de74495ca4.png"},{"id":92011521,"identity":"8b05c68b-3096-4124-889e-f728d7e6d2c7","added_by":"auto","created_at":"2025-09-23 15:45:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1396692,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7315296/v1/058b1b9b-7d24-4f8b-bf4a-bf91c0b562fc.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Trusting AI Too Much? Psychological Predictors of Overtrust and the Mitigating Role of AI Literacy","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eAs of the 2025, the advancement of artificial intelligence (AI) has brought about remarkable changes in the field of education, with generative AI such as ChatGPT leading innovative transformations in this area (Bhullar et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Jo, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). ChatGPT, a generative AI chatbot developed by OpenAI, utilizes natural language processing technology and large-scale language models (LLMs) to provide human-like conversational experiences (Bettayeb et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). ChatGPT is widely used by university students around the world for various purposes, including creative content generation, question answering, and language translation (Jo, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Li et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Alzain et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eChatGPT offers various benefits in the educational environment, such as personalized learning support and the promotion of critical thinking (Agbong-Coates, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Mai et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Students can receive assistance in their learning processes through close interaction with ChatGPT, and they can also receive real-time feedback from ChatGPT as a virtual teacher (Sok \u0026amp; Heng, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These features encourage student engagement and ensure the self-directed learning (Bhullar et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eHowever, there are several threats to the successful integration of ChatGPT in education (Hasanein \u0026amp; Sobaih, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Kovari, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Data privacy concerns, students\u0026rsquo; AI anxiety, fairness of academic outcomes derived from ChatGPT, and overtrust in AI are obstacles that can hinder successful learning (Bettayeb et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Sok \u0026amp; Heng, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Jo, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In particular, the phenomenon of AI overtrust, known as the AI trust paradox, refers to excessive trust on AI beyond its actual capabilities (Ullrich et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). It can cause direct difficulties in the learning and teaching processes for both students and teachers (Pithers \u0026amp; Soden, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2000\u003c/span\u003e). Overtrust in AI can be understood as a cognitive bias, where users exhibit misplaced confidence in AI outputs despite uncertain validity (Li et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). It can result not only in poor decision-making and decreased performance but also even in serious safety incidents (Parasuraman \u0026amp; Manzey, 2010). Moreover, overtrust in AI diminishes human critical engagement, thereby negatively impacting both collaborative outcomes with AI and ethical accountability (Ullrich et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Thus, overtrust in AI constitutes a significant risk factor at both individual and societal levels.\u003c/p\u003e\u003cp\u003eIt is essential for learners to develop AI literacy in the educational field, as it helps them integrate critical thinking and develop an ethical awareness of AI (Laitinen et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Fostering AI literacy in students is crucial to addressing significant AI-related side effects, such as overtrust in AI (Mohebi, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). While prior studies have consistently emphasized the necessity of enhancing ethical and analytical use of AI through AI literacy, empirical research examining the practical role of AI literacy in addressing of AI overtrust remains limited. In particular, although existing research has primarily focused on the risks of general machine overtrust (Lee \u0026amp; See, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Kaur et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ullrich et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), psychological research that identifies the antecedents of AI overtrust remains scarce. While there is some research on the psychological factors influencing AI dependency (Ye et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Acosta-Enriquez et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), little attention has been paid to investigating the psychological antecedents of AI overtrust. Therefore, this study contributes to understanding the psychological mechanisms underlying AI overtrust and examines whether AI literacy functions as a cognitive and behavioral buffer against its adverse consequences.\u003c/p\u003e"},{"header":"2. Literature review","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 AI overtrust and Learner Traits: Self-Efficacy and Motivation\u003c/h2\u003e\u003cp\u003eAI overtrust is considered a psychological state in which an individual\u0026rsquo;s level of trust in AI exceeds its actual capabilities (Aroyo et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Ullrich et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Such overtrust increases the likelihood of misuse of AI by underestimating the potential risks posed by AI (Aroyo et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). This can ultimately lead to excessive trust and reliance on AI (Klingbeil et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAccording to Social Cognitive Theory (SCT) (Bandura, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1997\u003c/span\u003e), human behavior results from the interaction of cognitive characteristics, behaviors, and environmental factors. Based on this theory, self-efficacy is one of the factors that can influence human overtrust in AI. Self-efficacy refers to individuals\u0026rsquo; beliefs in their own ability to perform tasks successfully (Bandura, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1997\u003c/span\u003e). Individuals with low self-efficacy are more likely to exhibit automation bias on machines and AI due to a lack of confidence in their own judgment (Patton, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Lee \u0026amp; Moray, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e1994\u003c/span\u003e). Consequently, lower self-efficacy is associated with greater AI overtrust (Wagner et al., 2018).\u003c/p\u003e\u003cp\u003e\u003cem\u003eH1: Students with lower self-efficacy will exhibit greater AI overtrust.\u003c/em\u003e\u003c/p\u003e\u003cp\u003eMeanwhile, Self-Determination Theory (SDT) posits that an individual's behavior is regulated according to the level of self-determination. Self-determination is composed of basic psychological needs: competence, autonomy, and relatedness (Deci \u0026amp; Ryan, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1985\u003c/span\u003e). According to Self-Determination Theory, when self-determination is satisfied, extrinsic motivation is initially driven by external incentives (Deci \u0026amp; Ryan, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1985\u003c/span\u003e), and subsequently develops into intrinsic motivation, which stems from interest in the activity itself (Ryan \u0026amp; Deci, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Roca \u0026amp; Gagn\u0026eacute;, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eFor individuals with strong intrinsic motivation, the intervention of autonomy-based cognitive forcing functions can reduce overtrust in AI (Bu\u0026ccedil;inca et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Klingbeil et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In contrast, those primarily driven by extrinsic motivation may be more prone to overtrusting AI due to a lack of critical appraisal of the AI technology (Gregor \u0026amp; Benbasat, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Eisbach et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cem\u003eH2: Students with higher intrinsic motivation will exhibit lower AI overtrust.\u003c/em\u003e\u003c/p\u003e\u003cp\u003e\u003cem\u003eH3: Students with higher extrinsic motivation will exhibit greater AI overtrust.\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 AI literacy as mediator\u003c/h2\u003e\u003cp\u003eAI literacy refers not only to the technical ability to operate AI but also to the comprehensive capacity to critically understand them from an ethical perspective (Čern\u0026yacute;, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). While literacy in traditional education primarily referred to basic skills such as reading, writing, and arithmetic (Čern\u0026yacute;, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Yim \u0026amp; Su, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), in contemporary society, the concept has expanded to encompass a broad understanding of and ability to navigate digital and intelligent information environments (Yim \u0026amp; Su, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Kong, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAI literacy comprises three key subcomponents. First, AI critical thinking involves the recognition of the limitations and potential errors of AI and the ability to avoid blindly accepting their outputs (Kong, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Dwivedi et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Second, AI ethics refers to an individual\u0026rsquo;s capacity to reflect on the social and ethical implications of AI, including concerns about data bias, privacy violations, and source reliability (Ng et al., 2022). Third, AI operation skill denotes the practical and technical proficiency in effectively operating AI tools and systems\u0026mdash;an essential competency for addressing digital inequality and improving access to information (Ng et al., 2022; Kong, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eA lack of AI literacy\u0026mdash;specifically, the inability to recognize the limitations and biases of AI and to critically reflect on its potential ethical implications for society\u0026mdash;can lead to overtrust in AI. Furthermore, insufficient understanding of or limited ability to use AI may also result in excessive reliance on and overtrust in AI (Kong, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Dwivedi et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Lee \u0026amp; See, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn general, high self-efficacy among learners is associated with ethical behavior (Stenmark et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and plays a critical role in the development of critical thinking skills (Dehghani et al., 2011). In particular, learners with a high level of autonomy\u0026mdash;a key driver of intrinsic motivation\u0026mdash;tend to adopt a critical perspective on AI through reflective cognition, thereby enhancing their AI literacy (He \u0026amp; Li, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Shen \u0026amp; Cui, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In contrast, while extrinsic motivation may be effective for achieving short-term goals or task performance, it has limitations in fostering a deeper understanding of AI technologies, ethical reasoning, and epistemological reflection (Artemova, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cem\u003eH4: AI literacy mediates the relationship between self-efficacy and AI overtrust.\u003c/em\u003e\u003c/p\u003e\u003cp\u003e\u003cem\u003eH5: AI literacy mediates the relationship between intrinsic motivation and AI overtrust.\u003c/em\u003e\u003c/p\u003e\u003cp\u003e\u003cem\u003eH6: AI literacy mediates the relationship between extrinsic motivation and AI overtrust.\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Performance expectation as mediator\u003c/h2\u003e\u003cp\u003ePerformance expectation refers to the degree to which an individual believes that using AI technology will enhance their performance (Venkatesh et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Dwivedi et al., 2019). It is a key factor that significantly influences the intention to use technology (Venkatesh et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), and a high level of performance expectation may lead to strong trust in AI technology (Harbarth et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Psychological factors that affect learners\u0026rsquo; performance expectations include self-efficacy and motivation. For instance, the more learners believe they can successfully perform academic tasks, the more strongly they believe in their ability to achieve desired learning outcomes (Lent et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Based on this theoretical background, the present study hypothesizes that the overtrust on AI stemming from learners\u0026rsquo; psychological factors is not only mediated by performance expectation but also exerts a sequential mediating effect in conjunction with AI literacy.\u003c/p\u003e\u003cp\u003eH7: Performance expectation will mediate the relationship between self-efficacy and AI overtrust.\u003c/p\u003e\u003cp\u003eH8: Performance expectation will mediate the relationship between intrinsic motivation and AI overtrust.\u003c/p\u003e\u003cp\u003eH9: Performance expectation will mediate the relationship between extrinsic motivation and AI overtrust.\u003c/p\u003e\u003cp\u003eH10: Performance expectation and AI literacy will sequentially mediate the relationship between learners\u0026rsquo; psychological traits and AI overtrust.\u003c/p\u003e\u003cp\u003eH10.1: Performance expectation and AI literacy will sequentially mediate the relationship between learners\u0026rsquo; intrinsic motivation and AI overtrust.\u003c/p\u003e\u003cp\u003eH10.2: Performance expectation and AI literacy will sequentially mediate the relationship between learners\u0026rsquo; extrinsic motivation and AI overtrust.\u003c/p\u003e\u003cp\u003eH10.3: Performance expectation and AI literacy will sequentially mediate the relationship between learners\u0026rsquo; self-efficacy and AI overtrust.\u003c/p\u003e\u003cp\u003eThe research model summarizing these hypothesized relationships is presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Materials and methods","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Participants\u003c/h2\u003e\u003cp\u003eThis study collected data through an online survey conducted in Mar 2025 via Embrain, a well-established online survey platform in South Korea. The study received prior approval from the institutional review board of our university. The participants consisted of currently enrolled university students in South Korea, including undergraduate, master\u0026rsquo;s, and doctoral students. Selection criteria required that all participants had prior experience using ChatGPT for academic-related purposes, such as completing coursework, studying, or conducting research. A total of 300 students participated in the survey. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the demographic characteristics of the participants.\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\u003eDemographic Information\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=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eFrequency\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePercent\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003e\u003cb\u003eSex\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=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e150\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e50.0\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=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e150\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e50.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e\u003cp\u003e\u003cb\u003eAge Group\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e19\u0026thinsp;~\u0026thinsp;24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e174\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e58.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25\u0026thinsp;~\u0026thinsp;29\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e97\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e32.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e30\u0026thinsp;~\u0026thinsp;34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e35~\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e\u003cp\u003e\u003cb\u003eDegree\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eUndergraduate Program\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e238\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e79.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMaster\u0026rsquo;s Program\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e11.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDoctoral Program\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e9.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e\u003cp\u003e\u003cb\u003eMajor\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHumanities\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e54\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e18.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSocial Sciences\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e68\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e22.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEngineering\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e91\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\u003eNatural Sciences\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e43\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e14.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eArts and Physical Education\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOthers\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e23\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eChatGPT Usage Frequency\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDaily\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e57\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e19.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4\u0026ndash;6 times a week\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e20.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1\u0026ndash;3 times a week\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e98\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e32.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1\u0026ndash;3 times a month\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e23.0\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAlmost never use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.3\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Measurements\u003c/h2\u003e\u003cdiv id=\"Sec9\" class=\"Section3\"\u003e\u003ch2\u003e3.2.1 Academic self-efficacy (Cronbach\u0026rsquo;s α\u0026thinsp;=\u0026thinsp;0.79)\u003c/h2\u003e\u003cp\u003eTo measure students\u0026rsquo; academic self-efficacy, the academic self-efficacy questionnaire developed by Nielsen et al. (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) was used. This questionnaire includes statements such as: \"I believe that with sufficient effort throughout the semester, I can solve academic difficulties by myself,\" \"I believe that with enough study, I can understand complex concepts,\" and \"I believe that even as the assignment deadline approaches, I can manage stress and meet the deadline.\" This instrument has been widely used in previous studies (Akanni \u0026amp; Oduaran, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Hitches et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Additionally, Naiseh et al. (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) explored the impact of academic self-efficacy on AI usage, showing that individuals with higher self-efficacy are more confident in using AI tools effectively for academic tasks.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section3\"\u003e\u003ch2\u003e3.2.2 Motivation (intrinsic motivation Cronbach\u0026rsquo;s α\u0026thinsp;=\u0026thinsp;0.75, extrinsic motivation Cronbach\u0026rsquo;s α\u0026thinsp;=\u0026thinsp;0.58)\u003c/h2\u003e\u003cp\u003eLearners\u0026rsquo; academic motivation was assessed using items adapted from the Self-Determination Theory (Deci \u0026amp; Ryan, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e1985\u003c/span\u003e), distinguishing between intrinsic (e.g., curiosity-driven learning) and extrinsic (e.g., grade-oriented effort) motivations. Items reflected both autonomy-driven and performance-driven attitudes towards learning, as suggested by prior studies integrating motivational constructs into technology acceptance (Nikou \u0026amp; Economides, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Lee et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2005\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\u003ch2\u003e3.2.3 Performance expectations (Cronbach\u0026rsquo;s α\u0026thinsp;=\u0026thinsp;0.82)\u003c/h2\u003e\u003cp\u003ePerformance expectations were measured using three items adapted from Aronson and Carlsmith (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e1962\u003c/span\u003e). These included statements such as \u0026ldquo;I think using ChatGPT will help me solve problems more easily when studying,\u0026rdquo; \u0026ldquo;I believe ChatGPT will provide helpful information for my learning,\u0026rdquo; and \u0026ldquo;I think using ChatGPT will save me time when studying.\u0026rdquo; The reliability of this scale has been previously confirmed (Nan et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003ea). Additionally, Venkatesh et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2003\u003c/span\u003e) explored the impact of performance expectations on attitudes toward technology in the context of UTAUT, highlighting that performance expectations significantly influence users' attitudes toward technology use.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\u003ch2\u003e3.2.4 AI literacy (Cronbach\u0026rsquo;s α\u0026thinsp;=\u0026thinsp;0.77)\u003c/h2\u003e\u003cp\u003eAI literacy was measured using items developed from prior research (e.g., Ng et al., 2012; Imjai et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Lin et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), capturing participants\u0026rsquo; understanding of AI fundamentals, practical AI proficiency, and awareness of ethical implications. The items covered cognitive and metacognitive aspects (Kong, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), including the ability to critically analyze AI-generated information, assess data sources, and adapt AI tools effectively in learning contexts. Participants were also asked to evaluate AI\u0026rsquo;s limitations, data use transparency, and social accountability, reflecting contemporary AI ethics frameworks.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section3\"\u003e\u003ch2\u003e3.2.5 AI overtrust (Cronbach\u0026rsquo;s α\u0026thinsp;=\u0026thinsp;0.85)\u003c/h2\u003e\u003cp\u003eTo assess the degree of overtrust on AI ChatGPT in academic decision-making contexts, an AI overtrust scale was developed based on prior research (e.g., Ullrich et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Kaur et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Klingbeil et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The scale comprised three items, such as \u0026ldquo;I believe that ChatGPT provides more accurate judgments than human instructors,\u0026rdquo; measuring users' tendency to overestimate the reliability and usefulness of AI-generated information. This construct reflects miscalibrated trust, where the perceived capabilities of AI exceed its actual performance in learning situations (Lee \u0026amp; See, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). All items were rated on a 5-point Likert scale (1\u0026thinsp;=\u0026thinsp;strongly disagree to 5\u0026thinsp;=\u0026thinsp;strongly agree).\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Data Analysis Strategy\u003c/h2\u003e\u003cp\u003eStatistical analysis was performed using IBM SPSS Statistics version 27. To test the mediation effects, the PROCESS macro version 4.2 (Hayes, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) was utilized, specifically using Model 4 to analyze the mediating role of performance expectation and AI literacy in the relationship between learner characteristics such as academic self-efficacy and academic motivation, and AI overtrust. The indirect effects were tested using 5,000 bootstrapping iterations with a 95% bias-corrected confidence interval. If the confidence interval of the indirect effect did not contain 0, the mediation effect was considered statistically significant (Preacher \u0026amp; Hayes, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). All variables underwent mean-centering to minimize multicollinearity issues. Prior to the analysis, the basic assumptions, including linearity and homoscedasticity, were checked, and all assumptions were met. Additionally, a preliminary review of the relationships between variables and multicollinearity indicated that no multicollinearity was present.\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Results","content":"\u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eDescriptive statistics and correlation coefficients\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eAOT\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eAILit\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.230**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.360**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.216**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAILit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.471**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.457**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.01\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.47\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\"\u003eNote: AOT, AI overtrust; SE, Self efficacy; PE, Performance Expectation; AILit, AI literacy; Confidence Interval; *p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, *** p\u0026thinsp;=\u0026thinsp;0.000.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eRegression Coefficients for the Serial Mediation of Self-Efficacy on AI overtrust\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eDependent variable\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eIndependent variable\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSE (HC3)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003et\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.2261\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0695\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.34**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.0952, .3690]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eAILit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.3823\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0415\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6.82***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.2016, .3651]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.3770\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0509\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.35***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.1720, .3724]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eAOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.4563\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0794\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e8.25***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.4994, .8121]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAILit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026ndash;.2847\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1272\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026ndash;4.45***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[\u0026ndash;.8170, \u0026ndash;.3163]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\"\u003eNote: PE, Performance Expectation; SE, Self efficacy; AILit, AI literacy; AOT, AI overtrust; \u0026beta;, Standardized regression coefficient; SE(HC3), Heteroscedasticity-consistent standard error (HC3); t, t-statistic; 95% CI, 95% Confidence Interval; *p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, *** p\u0026thinsp;=\u0026thinsp;0.000.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable id=\"Tab4\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eIndirect and Contrast Effects of Self-Efficacy on AI overtrust\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eIndirect Path\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eEffect\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eBoot SE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd1: SE \u0026rarr; PE \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1522\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0481\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.0648, .2536]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd2: SE \u0026rarr; AILit \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026ndash;.1606\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0423\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[\u0026ndash;.2492, \u0026ndash;.0838]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd3: SE \u0026rarr; PE \u0026rarr; AILit \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026ndash;.0358\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0145\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[\u0026ndash;.0691, \u0026ndash;.0131]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colspan=\"4\"\u003e\n \u003cp\u003e\u003cstrong\u003eIndirect Effect Contrasts\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC1: Ind1 \u0026ndash; Ind2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.3128\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0700\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.1843, .4605]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC2: Ind1 \u0026ndash; Ind3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1880\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0591\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.0804, .3123]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC3: Ind2 \u0026ndash; Ind3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026ndash;.1248\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0379\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[\u0026ndash;.2039, \u0026ndash;.0567]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\"\u003eNote: PE, Performance Expectation; SE, Self efficacy; AILit, AI literacy; AOT, AI overtrust; \u0026beta;, Standardized regression coefficient; Effect\u0026thinsp;=\u0026thinsp;Indirect effect estimate; Boot SE\u0026thinsp;=\u0026thinsp;Bootstrapped standard error of the indirect effect; 95% CI, 95% Confidence Interval\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable id=\"Tab5\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eSummary of Effects of Self-Efficacy on AI overtrust\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePath (SE \u0026rarr; AOT)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eEffect\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTotal Effect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.2971\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDirect Effect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.3413\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.1674, .5152]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndirect Effect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026ndash;.0442\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[\u0026ndash;.1494, .0696]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable id=\"Tab6\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eRegression Coefficients for the Serial Mediation of Motivation on AI overtrust\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colspan=\"6\"\u003e\n \u003cp\u003ePanel A. Intrinsic Motivation (IM)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDependent variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndependent variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSE (HC3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003et\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1304\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0539\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.42*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.0242, .2365]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eAILit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0759\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0348\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2.18*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.0073, .1444]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.3194\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0552\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5.79***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.2108, .4281]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eAOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePerEx\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.6407\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0810\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7.91***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.4813, .8000]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAILit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026minus;\u0026thinsp;.4033\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1312\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-3.07**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.6616, \u0026minus;\u0026thinsp;.1451]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colspan=\"6\"\u003e\n \u003cp\u003ePanel B. Extrinsic Motivation (EM)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDependent variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndependent variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSE (HC3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003et\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.3147\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0605\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5.21***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.1957, .4337]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eAILit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0298\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0495\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.0676, .1271]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.3239\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0619\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5.23***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.2021, .4457]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eAOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePerEx\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.5715\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0823\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6.94***\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.4095, .7334]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAILit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026minus;\u0026thinsp;.3742\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1277\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-2.93**\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.6256, \u0026minus;\u0026thinsp;.1229]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\"\u003eNote: PE, Performance Expectation; IM, Intrinsic Motivation; EM, Extrinsic Motivation; AILit, AI literacy; AOT, AI overtrust; \u0026beta;, Standardized regression coefficient; SE(HC3), Heteroscedasticity-consistent standard error (HC3); t, t-statistic; 95% CI, 95% Confidence Interval; *p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, *** p\u0026thinsp;=\u0026thinsp;0.000.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable id=\"Tab7\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eIndirect and Contrast Effects of Motivation on AI overtrust\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colspan=\"4\"\u003e\n \u003cp\u003ePanel A. Intrinsic Motivation (IM)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndirect Path\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEffect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBoot SE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd1: IM \u0026rarr; PE \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0835\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0352\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.0178, .1538]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd2: IM \u0026rarr; AILit \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026minus;\u0026thinsp;.0306\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0173\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.0701, \u0026minus;\u0026thinsp;.0031]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd3: IM \u0026rarr; PE \u0026rarr; AILit \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026minus;\u0026thinsp;.0168\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0096\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.0393, \u0026minus;\u0026thinsp;.0024]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colspan=\"4\"\u003e\n \u003cp\u003e\u003cstrong\u003eIndirect Effect Contrasts\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC1: Ind1 \u0026ndash; Ind2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1141\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0389\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.0444, .1957]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC2: Ind1 \u0026ndash; Ind3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0429\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.0215, .1882]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC3: Ind2 \u0026ndash; Ind3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026minus;\u0026thinsp;.0138\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0183\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.0533, .0194]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colspan=\"4\"\u003e\n \u003cp\u003ePanel B. Extrinsic Motivation (EM)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndirect Path\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEffect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBoot SE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd1: EM \u0026rarr; PE \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1799\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0408\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.1026, .2621]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd2: EM \u0026rarr; AILit \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026minus;\u0026thinsp;.0111\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0193\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.0527, .0262]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInd3: EM \u0026rarr; PE \u0026rarr; AILit \u0026rarr; AOT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026minus;\u0026thinsp;.0382\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0161\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.0748, \u0026minus;\u0026thinsp;.0130]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colspan=\"4\"\u003e\n \u003cp\u003e\u003cstrong\u003eIndirect Effect Contrasts\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC1: Ind1 \u0026ndash; Ind2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.1910\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0417\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.1111, .2735]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC2: Ind1 \u0026ndash; Ind3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.2180\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0506\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[.1236, .3220]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC3: Ind2 \u0026ndash; Ind3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0270\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e.0275\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-.0184, .0904]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\"\u003eNote: PE, Performance Expectation; IM, Intrinsic Motivation; EM, Extrinsic Motivation; AILit, AI literacy; AOT, AI overtrust; \u0026beta;, Standardized regression coefficient; Effect\u0026thinsp;=\u0026thinsp;Indirect effect estimate; Boot SE\u0026thinsp;=\u0026thinsp;Bootstrapped standard error of the indirect effect; 95% CI, 95% Confidence Interval\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable id=\"Tab8\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 8\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eSummary of Effects of Motivation on AI overtrust\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eEffect Type\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eIntrinsic Motivation\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e95% CI (Intrinsic)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eExtrinsic Motivation\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e95% CI (Extrinsic)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTotal Effect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.2184\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.4094\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDirect Effect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.1823\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[0.0544, 0.3101]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.2788\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[0.1304, 0.4272]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndirect Effect\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.0361\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[-0.0342, 0.1024]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.1306\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e[0.0459, 0.2138]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003e4.1 Descriptive Statistics and Correlations\u003c/h2\u003e\n \u003cp\u003eDescriptive statistics and correlation coefficients for the study variables are presented in Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e. AI overtrust (AOT) was positively correlated with self-efficacy (SE) (r\u0026thinsp;=\u0026thinsp;.230, p\u0026thinsp;\u0026lt;\u0026thinsp;.001) and performance expectation (PE) (r\u0026thinsp;=\u0026thinsp;.360, p\u0026thinsp;\u0026lt;\u0026thinsp;.001), while AI literacy (AILit) was not significantly correlated with AOT. Table \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e displays the regression coefficients for the serial mediation model of SE on AOT. SE had a significant positive effect on PE (\u0026beta;\u0026thinsp;=\u0026thinsp;.2261, p\u0026thinsp;\u0026lt;\u0026thinsp;.001) and AILit (\u0026beta;\u0026thinsp;=\u0026thinsp;.3823, p\u0026thinsp;=\u0026thinsp;.000), and PE also had a significant effect on AILit (\u0026beta;\u0026thinsp;=\u0026thinsp;.3770, p\u0026thinsp;=\u0026thinsp;.000). In turn, AILit had a significant negative effect on AOT (\u0026beta; = \u0026ndash;.2847, p\u0026thinsp;=\u0026thinsp;.000), while PE had a significant positive direct effect on AOT (\u0026beta;\u0026thinsp;=\u0026thinsp;.4563, p\u0026thinsp;=\u0026thinsp;.000). These results suggest that higher SE increases PE and AILit, with AILit reducing AOT and PE directly contributing to it.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003e4.2 Direct Effects of Psychological Predictors on AI overtrust\u003c/h2\u003e\n \u003cp\u003eThe analysis of the direct effects of SE and motivation variables (intrinsic motivation, extrinsic motivation) on AOT revealed that learners\u0026rsquo; SE had a significant positive impact on AOT (\u0026beta;\u0026thinsp;=\u0026thinsp;.3413, p\u0026thinsp;\u0026lt;\u0026thinsp;.01) (Table \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e), which contradicts hypothesis H1. This result also diverges from prior research (Patton, \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zhang et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e; Lee \u0026amp; Moray, \u003cspan class=\"CitationRef\"\u003e1994\u003c/span\u003e), which suggested that lower SE is associated with higher levels of AOT. Intrinsic motivation (IM) had a significant positive effect on AOT (\u0026beta;\u0026thinsp;=\u0026thinsp;.1823, p\u0026thinsp;\u0026lt;\u0026thinsp;.01) (Table \u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003e), contradicting hypothesis H2. Hypothesis H2 predicted that students with higher IM would exhibit lower levels of AOT; however, the actual findings revealed the opposite. Similarly, extrinsic motivation (EM) also showed a significant positive effect on AOT (\u0026beta;\u0026thinsp;=\u0026thinsp;.2788, p\u0026thinsp;\u0026lt;\u0026thinsp;.001) (Table \u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003e), supporting hypothesis H3. This indicates that students who are motivated by external rewards tend to have higher expectations regarding the functionality and results of AI, which in turn leads to increased AOT. The structural model for these relationships is shown in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e, where both IM and EM have significant positive effects on PE, which in turn influences AOT directly and indirectly through AILit.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003ch2\u003e4.3 Mediating Role of Performance Expectation and AI Literacy\u003c/h2\u003e\n \u003cp\u003eThe structural model of SE\u0026rsquo;s effects on AOT via PE and AILit is shown in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e. As illustrated, SE positively affected both PE and AILit, PE positively affected AOT, and AILit negatively affected AOT. Next, we examined the direct and sequential mediating effects of PE and AILit. The indirect path from SE through PE to AOT showed a significant positive mediating effect (\u0026beta;\u0026thinsp;=\u0026thinsp;.1522, Boot CI = [.0648, .2536]) (Table \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e), supporting H7. In contrast, the path from SE \u0026rarr; AILit \u0026rarr; AOT showed a significant negative mediating effect (\u0026beta; = \u0026ndash;.1606, CI = [\u0026ndash;.2492, \u0026ndash;.0838]) (Table \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e), supporting H4 and suggesting that AILit can suppress AOT. IM also had a positive indirect effect on AOT through PE (\u0026beta;\u0026thinsp;=\u0026thinsp;.0835, CI = [.0178, .1538]) (Table \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e), supporting H8, while the path via AILit showed a significant negative effect (\u0026beta; = \u0026ndash;.0306, CI = [\u0026ndash;.0701, \u0026ndash;.0031]) (Table \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e), supporting H5. For EM, the path via PE showed a significant positive indirect effect (\u0026beta;\u0026thinsp;=\u0026thinsp;.1799, CI = [.1026, .2621]) (Table \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e), supporting H9. The detailed regression coefficients for these serial mediation models of IM and EM are presented in Table \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e. The complex path from PE to AILit and then to critical thinking showed a significant negative indirect effect (\u0026beta; = \u0026ndash;.0382, CI = [\u0026ndash;.0748, \u0026ndash;.0130]) (Table \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e), supporting H6. However, the single path via only AILit (\u0026beta; = \u0026ndash;.0111) was not statistically significant, indicating that the direct path of H6 was not supported. Overall, these results suggest that SE and IM form pathways through AILit that critically mediate AOT, while EM does not show a significant mediating effect when AILit is considered alone.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\n \u003ch2\u003e4.4 Serial Mediation Effects: Performance expectation and AI Literacy\u003c/h2\u003e\n \u003cp\u003eThe analysis of the sequential mediating effects of PE and AILit revealed significant indirect paths for all psychological independent variables. For SE, a significant negative indirect effect was found in the path from SE through PE and AILit to AOT (\u0026beta; = \u0026ndash;.0358, CI = [\u0026ndash;.0691, \u0026ndash;.0131]) (Table \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e). This suggests that high SE increases PE, which leads to improved AILit, ultimately enhancing critical judgment toward AI and suppressing overtrust (supporting H10.3). A similar trend was observed for IM (\u0026beta; = \u0026ndash;.0168, CI = [\u0026ndash;.0393, \u0026ndash;.0024]) (Table \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e), supporting H10.1. For EM, the indirect effect in the sequential mediating path was the strongest (\u0026beta; = \u0026ndash;.0382, CI = [\u0026ndash;.0748, \u0026ndash;.0130]) (Table \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e), supporting H10.2. This result indicates that learners with EM, who prioritize external rewards, can also suppress overtrust in AI technologies if they sufficiently develop AILit.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"5. Discussion","content":"\u003cp\u003eThis study analyzed the impact of psychological factors such as self-efficacy, intrinsic motivation, and extrinsic motivation on AI overtrust, and examined the mediating and sequential mediating roles of performance expectation and AI literacy in this pathway. The key findings derived from this study are summarized in four main points.\u003c/p\u003e\u003cp\u003eFirst, higher self-efficacy was found to indirectly increase the level of AI overtrust. This result contradicts previous studies that argued lower self-efficacy leads to greater AI overtrust (Patton, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Lee \u0026amp; Moray, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Lee \u0026amp; See, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Lee and See (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) suggested that users with low confidence in their ability to use automated technologies tend to overtrust them. Similarly, Zhang et al. (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) found that learners with low self-efficacy are more likely to become dependent on AI. In contrast, the current analysis reveals an indirect pathway whereby high self-efficacy in learning leads to increased performance expectation, which in turn contributes to greater AI overtrust. This suggests that learners with strong self-efficacy may develop overly optimistic expectations about AI\u0026rsquo;s ability to enhance learning outcomes, leading to an inflated perception of AI capabilities. Such overestimation may foster excessive trust in AI (Lent et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eSecond, higher levels of intrinsic motivation were found to directly increase AI overtrust. This finding contrasts with prior studies that argued intrinsic motivation enhances cognitive engagement and critical thinking, thereby reducing overreliance on AI (Bu\u0026ccedil;inca et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Klingbeil et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The present result suggests that intrinsically motivated learners, while typically engaging in deeper cognitive processing, may develop emotional dependence on AI tools such as ChatGPT due to their repetitive and rewarding interaction patterns. This emotional attachment\u0026mdash;ironically rooted in high cognitive engagement\u0026mdash;can undermine critical thinking and foster uncritical acceptance of AI outputs (Yankouskaya et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThird, extrinsic motivation had a stronger direct and indirect effect on AI overtrust than intrinsic motivation. This may be attributed to the instrumental and performance-oriented learning context fostered by AI tools such as ChatGPT, which amplifies the influence of extrinsic motivation (Klingbeil et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Learners with high extrinsic motivation tend to view AI as a means to maximize learning efficiency, which elevates their performance expectation. This, in turn, leads to uncritical acceptance of AI outputs and fosters AI overtrust.\u003c/p\u003e\u003cp\u003eFourth, performance expectation and AI literacy function as key mediators in the formation of AI overtrust based on learners\u0026rsquo; psychological traits. The results show that when learners expect high academic efficiency from AI use, they tend to accept AI outputs uncritically. In contrast, AI literacy plays a crucial role in mitigating or blocking this path by enabling learners to understand the limitations of AI\u0026mdash;such as algorithmic bias, data opacity, and lack of accountability\u0026mdash;and to critically evaluate AI-generated information (Ng et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eNotably, while extrinsic motivation alone did not significantly reduce AI overtrust through AI literacy, it exhibited a strong negative effect when mediated sequentially by performance expectation and AI literacy. This suggests that extrinsically motivated learners may curb AI overtrust if their performance expectations lead to the development of AI literacy. These findings suggest that extrinsic motivation does not inherently lead to AI overtrust, but that cognitive mediation processes involving motivation, performance expectation, and AI literacy play a critical role in shaping this relationship.\u003c/p\u003e\u003cp\u003eOverall, this study highlights the risk that learners\u0026rsquo; psychological characteristics can lead to overtrust in AI and proposes new directions for AI literacy education and motivation-based learning strategies in AI-integrated environments.\u003c/p\u003e"},{"header":"6. Implications and Limitations","content":"\u003cp\u003eThe results offer several important implications for educational practice and policy. First, from a theoretical perspective, this study empirically demonstrates that AI literacy and performance expectation can function as a complex mediating structure explaining AI overtrust. While prior studies have focused on the impact of psychological factors on attitudes toward AI, this study offers new insights by clarifying how these factors lead to AI overtrust, emphasizing AI literacy as a key cognitive mediator in the process. From a practical standpoint, as the educational use of generative AI tools, including ChatGPT, continues to grow, this study suggests that the mere introduction of such technologies may result in unintended side effects, such as AI overtrust. Learners with high self-efficacy and strong intrinsic and extrinsic motivation are more likely to trust and use AI. However, when their AI literacy is low, this trust may easily develop into overtrust.\u003c/p\u003e\u003cp\u003eTherefore, it is essential to provide adequate pre-usage education tailored to learners\u0026rsquo; psychological characteristics. Such education should foster proper AI literacy while also enhancing awareness of AI\u0026rsquo;s limitations and potential errors, as well as learners\u0026rsquo; ability to critically assess AI outputs. Learners driven by extrinsic motivations, such as grades or financial rewards, are particularly vulnerable to overtrust. This highlights the need for a structured AI literacy curriculum targeting these groups.\u003c/p\u003e\u003cp\u003eThis study also presents several limitations. First, the sample was limited to university students in South Korea, which restricts the generalizability of the findings. Future research should include more diverse participant groups across age, cultural, and occupational backgrounds. Second, the cross-sectional design limits the ability to establish causal relationships among variables. Longitudinal studies are needed to explore how psychological factors and overtrust evolve with continued AI use. Third, all variables were measured through self-report questionnaires, which may be influenced by social desirability bias or subjective interpretation. Future studies should consider incorporating behavioral data or AI usage logs to improve data reliability.\u003c/p\u003e\u003cp\u003eIn conclusion, this study identifies the potential for learners\u0026rsquo; psychological characteristics to lead to AI overtrust in educational settings. It also empirically demonstrates that AI literacy plays a critical mediating role in this process. These findings offer foundational insights for future AI integration in education and for the development of related policy frameworks. Specifically, the study emphasizes that AI literacy education must extend beyond technical skills. It should incorporate critical thinking, an understanding of algorithmic transparency, and awareness of data ethics and social responsibility. These elements are essential and should be meaningfully integrated into educational practice.\u003c/p\u003e"},{"header":"7. Future Research Directions","content":"\u003cp\u003eThis study is significant in empirically identifying the psychological factors influencing AI overtrust and the mediating factors involved. However, to better understand these results in a more refined and expanded form, follow-up research from various perspectives is necessary. First, since this study was conducted as a cross-sectional study using data from a specific point in time, it is difficult to clearly identify changes over time or causal flows between the variables. Future research should use a longitudinal design to track how learners\u0026rsquo; AI usage experiences affect their self-efficacy and motivation, and how these changes, in turn, influence the level of AI overtrust over the long term.\u003c/p\u003e\u003cp\u003eAdditionally, the sample in this study was limited to university students in South Korea, meaning that cultural context, age, education level, and personality traits were not considered. Therefore, future studies should expand the research to include learner groups from diverse cultural backgrounds and analyze whether the impact of psychological factors on AI overtrust varies across cultures. Particularly in countries with different social perceptions of AI, technological accessibility, or educational environments, the results from this study may differ.\u003c/p\u003e\u003cp\u003eFurthermore, this study was based on self-reported surveys, which have the limitation of relying on subjective perceptions. In the future, it would be beneficial to collect a combination of quantitative and qualitative data, such as actual AI tool usage logs, learning performance indicators, and objective literacy assessment tools, to derive more accurate and comprehensive results. For example, analyzing the quality of texts written by learners using ChatGPT or examining judgment errors in the way tasks are solved could empirically confirm the impact of AI overtrust on the actual learning process.\u003c/p\u003e\u003cp\u003eLastly, this study analyzed the mediating effects focusing on performance expectation and AI literacy. However, future studies should include a broader range of cognitive and emotional factors, such as AI trust, the ability to assess the authenticity of information, and tolerance for ambiguity, to construct a more multi-layered explanatory model of AI overtrust. Moreover, research that analyzes the perceptions and usage of AI by not only learners but also teachers, educators, and policymakers is required to enhance the ecological understanding of educational AI usage.\u003c/p\u003e\u003cp\u003eIn conclusion, as AI technology rapidly spreads, research that goes beyond analyzing learner-centered psychological factors and encompasses various contexts, groups, and long-term effects will significantly contribute to the practical strategies for improving the quality of AI education and preventing overtrust.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eAI: \u0026nbsp; Artificial Intelligence\u003c/p\u003e\n\u003cp\u003eLLM: \u0026nbsp; Large Language Model\u003c/p\u003e\n\u003cp\u003eAOT: \u0026nbsp; AI Overtrust\u003c/p\u003e\n\u003cp\u003eSE: \u0026nbsp; Self-Efficacy\u003c/p\u003e\n\u003cp\u003eIM: \u0026nbsp; Intrinsic Motivation\u003c/p\u003e\n\u003cp\u003eEM: \u0026nbsp; Extrinsic Motivation\u003c/p\u003e\n\u003cp\u003ePE: \u0026nbsp; Performance Expectation\u003c/p\u003e\n\u003cp\u003eAILit: \u0026nbsp; AI Literacy\u003c/p\u003e\n\u003cp\u003eCI: \u0026nbsp; Confidence Interval\u003c/p\u003e\n\u003cp\u003eSE (HC3): \u0026nbsp; Standard Error (Heteroscedasticity-consistent type 3)\u003c/p\u003e\n\u003cp\u003eBoot SE: \u0026nbsp; Bootstrapped Standard Error\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors and Affiliations \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDepartment of Interaction Science, Department of Human-Artificial Intelligence Interaction, Sungkyunkwan University, 03063 Seoul, Republic of Korea \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eYoungsang Kim \u0026amp; Jang Hyun Kim\u003c/p\u003e\n\u003cp\u003eInstitute of Chinese Language and Culture Education, College of Chinese Language and Culture, Huaqiao University, Xiamen, China\u003c/p\u003e\n\u003cp\u003eShunan Zhang\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYoungsang Kim: conceptualization, data analysis, and original draft writing; \u0026nbsp;Shunan Zhang: conceptualization and supervision; \u0026nbsp;Jang Hyun Kim: conceptualization, supervision, funding acquisition, and manuscript review and editing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorresponding author:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrespondence to Shunan Zhang\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll procedures involving human participants in this study were conducted in accordance with the ethical standards of the Declaration of Helsinki. The study protocol was reviewed and approved by the Institutional Review Board of Sungkyunkwan University (IRB approval number: 2025-02-081). Written informed consent was obtained from all participants prior to their inclusion in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWritten informed consent for publication of anonymized data was obtained from all participants prior to their inclusion in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author declares no conflict of interest\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data and material used to derive the findings in this study are available from the corresponding author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAcosta-Enriquez, B. G., Ballesteros, M. A. A., Valle, M. D. L. A. G., Angaspilco, J. E. M., Lalup\u0026uacute;, J. D. R. A., Jaico, J. L. B., ... \u0026amp; Castillo, W. E. J. (2025). The mediating role of academic stress, critical thinking and performance expectations in the influence of academic self-efficacy on AI dependence: case study in college students. \u003cem\u003eComputers and Education: Artificial Intelligence\u003c/em\u003e\u003cem\u003e 8\u003c/em\u003e, 100381. https://doi.org/10.1016/j.caeai.2025.100381\u003c/li\u003e\n\u003cli\u003eAgbong-Coates, I. J. G. (2024). ChatGPT integration significantly boosts personalized learning outcomes: A Philippine study. \u003cem\u003eInternational journal of educational management and development studies, 5(2)\u003c/em\u003e, 165-186. https://doi.org/10.53378/353067\u003c/li\u003e\n\u003cli\u003eAkanni, A. A., \u0026amp; Oduaran, C. A. (2018). Perceived social support and life satisfaction among freshmen: Mediating roles of academic self-efficacy and academic adjustment. \u003cem\u003eJournal of Psychology in Africa, 28(2)\u003c/em\u003e, 89-93. https://doi.org/10.1080/14330237.2018.1454582\u003c/li\u003e\n\u003cli\u003eAlzain, E., Nagi, K. A., \u0026amp; Algobaei, F. (2024). The Quality of Google Translate and ChatGPT English to Arabic Translation: The Case of Scientific Text Translation. \u003cem\u003eIn Forum for Linguistic Studies\u003c/em\u003e\u003cem\u003e, 6(3)\u003c/em\u003e, pp. 837-849). https://doi.org/10.30564/fls.v6i3.6799\u003c/li\u003e\n\u003cli\u003eAronson, E., \u0026amp; Carlsmith, J. M. (1962). Performance expectation as a determinant of actual performance\u003cem\u003e. The Journal of Abnormal and Social Psychology, 65(3)\u003c/em\u003e, 178. https://doi.org/10.1037/h0042291\u003c/li\u003e\n\u003cli\u003eAroyo, A. M., De Bruyne, J., Dheu, O., Fosch-Villaronga, E., Gudkov, A., Hoch, H., ... \u0026amp; Tam\u0026ograve;-Larrieux, A. (2021). Overtrusting robots: Setting a research agenda to mitigate overtrust in automation. \u003cem\u003ePaladyn, Journal of Behavioral Robotics, 12(1)\u003c/em\u003e, 423-436. http://doi.org/10.1515/pjbr-2021-0029\u003c/li\u003e\n\u003cli\u003eArtemova, I. (2024). Bridging Motivation and AI in Education: An Activity Theory Perspective. \u003cem\u003eDigital Education Review, 45\u003c/em\u003e, 59-69. https://doi.org/ 10.1344/der.2024.45.59-67\u003c/li\u003e\n\u003cli\u003eBandura, A., \u0026amp; Wessels, S. (1997). \u003cem\u003eSelf-efficacy (pp. 4-6)\u003c/em\u003e. Cambridge: Cambridge University Press. \u003c/li\u003e\n\u003cli\u003eBettayeb, A. M., Abu Talib, M., Sobhe Altayasinah, A. Z., \u0026amp; Dakalbab, F. (2024). Exploring the impact of ChatGPT: conversational AI in education. \u003cem\u003eFrontiers in Education, 9\u003c/em\u003e, 1379796. https://doi.org/10.3389/feduc.2024.1379796.\u003c/li\u003e\n\u003cli\u003eBhullar, P. S., Joshi, M., \u0026amp; Chugh, R. (2024). ChatGPT in higher education-a synthesis of the literature and a future research agenda. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e\u003cem\u003e 29\u003c/em\u003e, 1-22. http://doi.org/10.1007/s10639-024-12723-x. \u003c/li\u003e\n\u003cli\u003eBu\u0026ccedil;inca, Z., Malaya, M. B., \u0026amp; Gajos, K. Z. (2021). To trust or to think: cognitive forcing functions can reduce overreliance on AI in AI-assisted decision-making. \u003cem\u003eProceedings of the ACM on Human-computer Interaction,5(CSCW1)\u003c/em\u003e, 1-21. http://doi.org/10.1145/3449287. \u003c/li\u003e\n\u003cli\u003eČern\u0026yacute;, M. (2024). University students\u0026rsquo; conceptualisation of AI literacy: Theory and empirical evidence. \u003cem\u003eSocial Sciences, 13(3)\u003c/em\u003e, 129. http://doi.org/10.3390/socsci13030129. \u003c/li\u003e\n\u003cli\u003eDeci, E. L., \u0026amp; Ryan, R. M. (1985). Conceptualizations of intrinsic motivation and self-determination\u003cem\u003e. Intrinsic motivation and self-determination in human behavior\u003c/em\u003e, 11-40. http://doi.org/10.1007/978-1-4899-2271-7_2. \u003c/li\u003e\n\u003cli\u003eDwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., ... \u0026amp; Wright, R. (2023). Opinion Paper: \u0026ldquo;So what if ChatGPT wrote it?\u0026rdquo; Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy. \u003cem\u003eInternational Journal of Information Management\u003c/em\u003e, 71, 102642. http://doi.org/10.1016/j.ijinfomgt.2023.102642. \u003c/li\u003e\n\u003cli\u003eEisbach, S., Langer, M., \u0026amp; Hertel, G. (2023). Optimizing human-AI collaboration: Effects of motivation and accuracy information in AI-supported decision-making. \u003cem\u003eComputers in Human Behavior: Artificial Humans,1(2),\u003c/em\u003e 100015. http://doi.org/10.1016/j.chbah.2023.100015. \u003c/li\u003e\n\u003cli\u003eGregor, S., \u0026amp; Benbasat, I. (1999). Explanations from intelligent systems: Theoretical foundations and implications for practice. \u003cem\u003eMIS quarterly\u003c/em\u003e, \u003cem\u003e23(4)\u003c/em\u003e, 497-530. http://doi.org/10.2307/249487. \u003c/li\u003e\n\u003cli\u003eHarbarth, L., G\u0026ouml;\u0026szlig;wein, E., Bodemer, D., \u0026amp; Schnaubert, L. (2025). (Over) trusting AI recommendations: How system and person variables affect dimensions of complacency. \u003cem\u003eInternational Journal of Human\u0026ndash;Computer Interaction, 41(1)\u003c/em\u003e, 391-410. http://doi.org/10.1080/10447318.2023.2301250. \u003c/li\u003e\n\u003cli\u003eHasanein, A. M., \u0026amp; Sobaih, A. E. E. (2023). Drivers and consequences of ChatGPT use in higher education: Key stakeholder perspectives. European journal of investigation in health, \u003cem\u003epsychology and education, 13(11),\u003c/em\u003e 2599-2614. http://doi.org/10.3390/ejihpe13110181. \u003c/li\u003e\n\u003cli\u003eHayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable mediation, moderation, and conditional process modeling.\u003c/li\u003e\n\u003cli\u003eHe, L., \u0026amp; Li, C. (2023). Continuance intention to use mobile learning for second language acquisition based on the technology acceptance model and self-determination theory. \u003cem\u003eFrontiers in Psychology, 14\u003c/em\u003e, 1185851. https://doi.org/10.3389/fpsyg.2023.1185851. \u003c/li\u003e\n\u003cli\u003eHitches, E., Woodcock, S., \u0026amp; Ehrich, J. (2022). Building self-efficacy without letting stress knock it down: Stress and academic self-efficacy of university students\u003cem\u003e. International Journal of Educational Research Open, 3\u003c/em\u003e, 100124. http://doi.org/10.1016/j.ijedro.2022.100124. \u003c/li\u003e\n\u003cli\u003eImjai, N., Yordudom, T., Yaacob, Z., Saad, N. H. M., \u0026amp; Aujirapongpan, S. (2025). Impact of AI literacy and adaptability on financial analyst skills among prospective Thai accountants: The role of critical thinking. \u003cem\u003eTechnological Forecasting and Social Change, 210\u003c/em\u003e, 123889. https://doi.org/10.1016/j.techfore.2024.123889. \u003c/li\u003e\n\u003cli\u003eJo, H. (2024). From concerns to benefits: a comprehensive study of ChatGPT usage in education. \u003cem\u003eInternational Journal of Educational Technology in Higher Education, 21(1)\u003c/em\u003e, 35. http://doi.org/10.1186/s41239-024-00471-4. \u003c/li\u003e\n\u003cli\u003eKaur, D., Uslu, S., Rittichier, K. J., \u0026amp; Durresi, A. (2022). Trustworthy artificial intelligence: a review. \u003cem\u003eACM computing surveys (CSUR), 55(2)\u003c/em\u003e, 1-38. https://doi.org/10.1145/3491209. \u003c/li\u003e\n\u003cli\u003eKlingbeil, A., Gr\u0026uuml;tzner, C., \u0026amp; Schreck, P. (2024). Trust and reliance on AI\u0026mdash;An experimental study on the extent and costs of overreliance on AI. \u003cem\u003eComputers in Human Behavior, 160\u003c/em\u003e, 108352. http://doi.org/10.1016/j.chb.2024.108352, \u003c/li\u003e\n\u003cli\u003eKong, S. C. (2014). Developing information literacy and critical thinking skills through domain knowledge learning in digital classrooms: An experience of practicing flipped classroom strategy. \u003cem\u003eComputers \u0026amp; education, 78\u003c/em\u003e, 160-173. http://doi.org/10.1016/j.compedu.2014.05.009. \u003c/li\u003e\n\u003cli\u003eKovari, A. (2025). Ethical use of ChatGPT in education\u0026mdash;Best practices to combat AI-induced plagiarism. In Frontiers in Education. \u003cem\u003eFrontiers Media SA\u003c/em\u003e\u003cem\u003e, 9\u003c/em\u003e, 1465704, https://doi.org/10.3389/feduc.2024.1465703. \u003c/li\u003e\n\u003cli\u003eLaitinen, S., Lepola, J., \u0026amp; Vauras, M. (2017). Early motivational orientation profiles and language comprehension skills: From preschool to Grade 3. \u003cem\u003eLearning and Individual Differences, 53\u003c/em\u003e, 69-78. \u003c/li\u003e\n\u003cli\u003eLee, J. D., \u0026amp; Moray, N. (1994). Trust, self-confidence, and operators\u0026rsquo; adaptation to automation. \u003cem\u003eInternational journal of human-computer studies, 40(1)\u003c/em\u003e, 153-184. https://doi.org/10.1006/ijhc.1994.1007. \u003c/li\u003e\n\u003cli\u003eLee, J. D., \u0026amp; See, K. A. (2004). Trust in automation: Designing for appropriate reliance. \u003cem\u003eHuman factors, 46(1),\u003c/em\u003e 50-80. https://doi.org/10.1518/hfes.46.1.50_30392. \u003c/li\u003e\n\u003cli\u003eLee, M. K., Cheung, C. M., \u0026amp; Chen, Z. (2005). Acceptance of Internet-based learning medium: the role of extrinsic and intrinsic motivation. \u003cem\u003eInformation \u0026amp; management, 42(8),\u003c/em\u003e 1095-1104. \u003c/li\u003e\n\u003cli\u003eLent, R. W., Sheu, H. B., Singley, D., Schmidt, J. A., Schmidt, L. C., \u0026amp; Gloster, C. S. (2008). Longitudinal relations of self-efficacy to outcome expectations, interests, and major choice goals in engineering students. \u003cem\u003eJournal of vocational behavior, 73(2)\u003c/em\u003e, 328-335. https://doi.org/10.1016/j.jvb.2008.07.005. \u003c/li\u003e\n\u003cli\u003eLi, Z., Wang, C., \u0026amp; Bonk, C. J. (2024). Exploring the Utility of ChatGPT for Self-Directed Online Language Learning. \u003cem\u003eOnline Learning, 28(3),\u003c/em\u003e 157-180. https://eric.ed.gov/?id=EJ1446866. \u003c/li\u003e\n\u003cli\u003eLin, X. F., Zhou, Y., Shen, W., Luo, G., Xian, X., \u0026amp; Pang, B. (2024). Modeling the structural relationships among Chinese secondary school students\u0026rsquo; computational thinking efficacy in learning AI, AI literacy, and approaches to learning AI. \u003cem\u003eEducation and Information Technologies, 29(5)\u003c/em\u003e, 6189-6215. http://doi.org/10.1007/s10639-023-12029-4. \u003c/li\u003e\n\u003cli\u003eMai, D. T. T., Da, C. V., \u0026amp; Hanh, N. V. (2024). The use of ChatGPT in teaching and learning: a systematic review through SWOT analysis approach. In Frontiers in Education. \u003cem\u003eFrontiers Media SA\u003c/em\u003e\u003cem\u003e, 9\u003c/em\u003e, 1327769, http://doi.org/10.3389/feduc.2024.1328769. \u003c/li\u003e\n\u003cli\u003eMohebi, L. (2024). Empowering learners with ChatGPT: insights from a systematic literature exploration. \u003cem\u003eDiscover Education, 3(1)\u003c/em\u003e, 36. http://doi.org/10.1007/s44217-024-00120-y. \u003c/li\u003e\n\u003cli\u003eNaiseh, M., Babiker, A., Al-Shakhsi, S., Cemiloglu, D., Al-Thani, D., Montag, C., \u0026amp; Ali, R. (2025). Attitudes Towards AI: The Interplay of Self-Efficacy, Well-Being, and Competency\u003cem\u003e. Journal of Technology in Behavioral Science,\u003c/em\u003e 1-14. http://doi.org/10.1007/s41347-025-00486-2. \u003c/li\u003e\n\u003cli\u003eNan, D., Lee, H., Kim, Y., \u0026amp; Kim, J. H. (2022). My video game console is so cool! A coolness theory-based model for intention to use video game consoles. \u003cem\u003eTechnological Forecasting and Social Change, 176\u003c/em\u003e, 121451. http://doi.org/10.1016/j.techfore.2021.121451. \u003c/li\u003e\n\u003cli\u003eNg, D. T. K., Leung, J. K. L., Chu, S. K. W., \u0026amp; Qiao, M. S. (2021). Conceptualizing AI literacy: An exploratory review. \u003cem\u003eComputers and Education: Artificial Intelligence,\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cem\u003e2\u003c/em\u003e, 100041. https://doi.org/10.1016/j.caeai.2021.100041. \u003c/li\u003e\n\u003cli\u003eNielsen, T., Dammeyer, J., Vang, M. L., \u0026amp; Makransky, G. (2018). Gender fairness in self-efficacy? A Rasch-based validity study of the General Academic Self-efficacy scale (GASE). \u003cem\u003eScandinavian Journal of Educational Research, 62(5)\u003c/em\u003e, 664-681. http://doi.org/10.1080/00313831.2017.1306796. \u003c/li\u003e\n\u003cli\u003eNikou, S. A., \u0026amp; Economides, A. A. (2017). Mobile-Based Assessment: Integrating acceptance and motivational factors into a combined model of Self-Determination Theory and Technology Acceptance. \u003cem\u003eComputers in Human Behavior, 68\u003c/em\u003e, 83-95. https://doi.org/10.1016/j.chb.2016.11.020Get rights and content.\u003c/li\u003e\n\u003cli\u003ePatton, C. E. (2023). The Influence of Trust, Self-Confidence and Task Difficulty on Automation Use (Doctoral dissertation, Colorado State University).\u003c/li\u003e\n\u003cli\u003ePithers, R. T., \u0026amp; Soden, R. (2000). Critical thinking in education: \u003cem\u003eA review. Educational research, 42(3)\u003c/em\u003e, 237-249. https://doi.org/10.1080/001318800440579\u003c/li\u003e\n\u003cli\u003ePreacher, K. J., \u0026amp; Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. \u003cem\u003eBehavior research methods, 40(3)\u003c/em\u003e, 879-891. http://doi.org/10.3758/BRM.40.3.879.\u003c/li\u003e\n\u003cli\u003eRoca, J. C., \u0026amp; Gagn\u0026eacute;, M. (2008). Understanding e-learning continuance intention in the workplace: A self-determination theory perspective. \u003cem\u003eComputers in human behavior, 24(4)\u003c/em\u003e, 1585-1604. https://doi.org/10.1016/j.chb.2007.06.001.\u003c/li\u003e\n\u003cli\u003eRyan, R. M., \u0026amp; Deci, E. L. (2000). Intrinsic and extrinsic motivations: Classic definitions and new directions. \u003cem\u003eContemporary educational psychology, 25(1)\u003c/em\u003e, 54-67. https://doi.org/10.1006/ceps.1999.1020.\u003c/li\u003e\n\u003cli\u003eShen, Y., \u0026amp; Cui, W. (2024). Perceived support and AI literacy: the mediating role of psychological needs satisfaction. \u003cem\u003eFrontiers in Psychology, 15\u003c/em\u003e, 1415248. https://doi.org/10.3389/fpsyg.2024.1415248.\u003c/li\u003e\n\u003cli\u003eSok, S., \u0026amp; Heng, K. (2023). ChatGPT for education and research: A review of benefits and risks. \u003cem\u003eCambodian Journal of Educational Research, 3(1)\u003c/em\u003e, 110-121. http://dx.doi.org/10.2139/ssrn.4378735\u003c/li\u003e\n\u003cli\u003eStenmark, C. K., Redfearn, R. A., \u0026amp; Kreitler, C. M. (2021). Self-efficacy and ethical decision-making. \u003cem\u003eEthics \u0026amp; Behavior, 31(5)\u003c/em\u003e, 301-320. https://doi.org/10.1080/10508422.2020.1776617.\u003c/li\u003e\n\u003cli\u003eUllrich, D., Butz, A., \u0026amp; Diefenbach, S. (2021). The development of overtrust: An empirical simulation and psychological analysis in the context of human\u0026ndash;robot interaction. \u003cem\u003eFrontiers in Robotics and AI, 8\u003c/em\u003e, 554578. http://doi.org/10.3389/frobt.2021.554578.\u003c/li\u003e\n\u003cli\u003eVenkatesh, V., Morris, M. G., Davis, G. B., \u0026amp; Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. \u003cem\u003eMIS quarterly, \u003c/em\u003e\u003cem\u003e27(3)\u003c/em\u003e, 425-478. http://doi.org/10.2307/30036540.\u003c/li\u003e\n\u003cli\u003eYankouskaya, A., Liebherr, M., \u0026amp; Ali, R. (2025). Can ChatGPT Be Addictive? A Call to Examine the Shift from Support to Dependence in AI Conversational Large Language Models. \u003cem\u003eHuman-Centric Intelligent Systems, \u003c/em\u003e\u003cem\u003e5\u003c/em\u003e, 77-89, https://doi.org/10.1007/s44230-025-00090-w\u003c/li\u003e\n\u003cli\u003eYe, J. H., Zhang, M., Nong, W., Wang, L., \u0026amp; Yang, X. (2025). The relationship between inert thinking and ChatGPT dependence: An I-PACE model perspective. \u003cem\u003eEducation and information technologies, 30(3)\u003c/em\u003e, 3885-3909. https://doi.org/10.1007/s10639-024-12966-8\u003c/li\u003e\n\u003cli\u003eYim, I. H. Y., \u0026amp; Su, J. (2024). Artificial intelligence (AI) learning tools in K-12 education: A scoping review. Journal of Computers in Education, 12(1), 93-131. https://doi.org/10.1007/s40692-023-00304-9\u003c/li\u003e\n\u003cli\u003eZhang, S., Che, S., Nan, D., \u0026amp; Kim, J. H. (2023). How does online social interaction promote students\u0026rsquo; continuous learning intentions? \u003cem\u003eFrontiers in Psychology\u003c/em\u003e, 14, 1098110. http://doi.org/10.3389/fpsyg.2023.1098110.\u003c/li\u003e\n\u003cli\u003eZhang, S., Zhao, X., Zhou, T., \u0026amp; Kim, J. H. (2024). Do you have AI dependency? The roles of academic self-efficacy, academic stress, and performance expectations on problematic AI usage behavior. \u003cem\u003eInternational Journal of Educational Technology in Higher Education, 21(1)\u003c/em\u003e. https://doi.org/10.1186/s41239-024-00467-0\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-psychology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"psyo","sideBox":"Learn more about [BMC Psychology](http://bmcpsychology.biomedcentral.com/)","snPcode":"","submissionUrl":"","title":"BMC Psychology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"AI overtrust, self-efficacy, intrinsic motivation, extrinsic motivation, AI literacy, performance expectation, structural equation modeling","lastPublishedDoi":"10.21203/rs.3.rs-7315296/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7315296/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eThe increasing adoption of AI tools such as ChatGPT in academic contexts has introduced a new psychological risk: overtrust in AI. However, the psychological foundations and boundary conditions of AI overtrust remain insufficiently examined. This study examined how self-efficacy, intrinsic motivation, and extrinsic motivation influence AI overtrust, and whether AI literacy can mitigate its negative consequences.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eA cross-sectional survey was conducted with 300 university students in South Korea who had prior academic experience using AI tools. The questionnaire measured academic self-efficacy, intrinsic and extrinsic motivation, AI literacy, and AI overtrust. Structural equation modeling was used to test direct and mediated effects among the variables, with bootstrapping employed to evaluate indirect pathways.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eSelf-efficacy and intrinsic motivation were positively associated with AI overtrust, suggesting that learners with higher confidence and autonomy exhibit overtrust in AI-generated information. Extrinsic motivation showed a weaker direct effect. AI literacy had a significant negative effect on AI overtrust and mediated the relationship between both self-efficacy and intrinsic motivation and overtrust. A sequential mediation model indicated that performance expectation and AI literacy together suppressed overtrust by strengthening reflective awareness.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eThese findings demonstrate that high-performing and intrinsically motivated learners are psychologically vulnerable to AI overtrust. Although high academic self-efficacy fosters autonomy in using AI tools, it also increases the risk of overtrust in AI. Targeted AI literacy education addresses this risk by enhancing learners\u0026rsquo; critical thinking, ethical awareness, and responsible AI use. AI literacy education should be adapted to learners\u0026rsquo; psychological profiles to effectively prevent AI overtrust and support responsible use.\u003c/p\u003e","manuscriptTitle":"Trusting AI Too Much? Psychological Predictors of Overtrust and the Mitigating Role of AI Literacy","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-23 15:29:50","doi":"10.21203/rs.3.rs-7315296/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-12-19T06:21:17+00:00","index":"","fulltext":""},{"type":"reviewerAgreed","content":"203775971965050749110051088483648285186","date":"2025-09-24T21:26:14+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-21T12:17:08+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"201671978825782520661975116782648662559","date":"2025-09-15T14:30:46+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-15T14:04:18+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"2056377821348091591927107606040391157","date":"2025-09-15T13:59:30+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-15T09:51:30+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-08-19T11:28:29+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-12T22:40:24+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-12T22:40:02+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Psychology","date":"2025-08-07T06:21:01+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-psychology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"psyo","sideBox":"Learn more about [BMC Psychology](http://bmcpsychology.biomedcentral.com/)","snPcode":"","submissionUrl":"","title":"BMC Psychology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"136d48ac-61a3-49d5-a0f3-a99940d45678","owner":[],"postedDate":"September 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-15T11:24:22+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-23 15:29:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7315296","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7315296","identity":"rs-7315296","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}