Abstract
Background
Artificial intelligence (AI) is reshaping healthcare from clinical care and diagnostics to operations and public health and has been accelerated by recent advances in large language models (LLMs), such as ChatGPT. Because healthcare directly concerns human life and well-being, the use of AI must be approached with care. As AI systems are ultimately designed, implemented, and trusted by people, identifying the factors that shape medical students’ attitudes toward AI is critical for safe and effective integration.
Methods
In this study, attitudes toward AI were measured with the General Attitudes toward AI Scale (GAAIS), and personality traits were measured with the Big Five Inventory-2 (BFI-2). Along with demographic information, all these data were collected via an online platform from five-year clinical medicine students in Beijing, Shanghai, and Chengdu. Correlation and linear regression were conducted.
Results
Analyses indicated that openness and agreeableness were associated with higher scores of positive attitudes toward AI, whereas conscientiousness was associated with lower scores of positive attitudes toward AI. For negative attitudes toward AI, openness and agreeableness were associated with its higher scores, whereas neuroticism was associated with its lower scores.
Conclusion
These results suggest that students with greater openness and agreeableness not only view AI more favorably but are also more tolerant of its limitations. In contrast, those high in conscientiousness report fewer positive views, and those high in neuroticism are less tolerant of AI shortcomings. To our knowledge, this is the first study to examine the associations between personality traits and attitudes toward AI in medical students in China, highlighting the need for targeted educational interventions that reflect diverse personality profiles.
Keywords: artificial intelligence, big five inventory, general attitudes toward AI scale, medical education, medical students
Introduction
Artificial intelligence (AI) is transforming healthcare, permeating clinical care, diagnostics, operations, and public health, with the development of AI in recent years, especially the breakthrough of large language models (LLMs) such as the ChatGPT. AI can increase efficiency, accuracy, and equity across different aspects of healthcare systems, such as computer vision, which marks lesions on images; predictive analytics, which anticipate disease outcomes; robotic assistance in procedures; and LLM tools, which streamline documentation and patient communication (1, 2). Since AI has been rapidly adopted in healthcare systems and is gaining increasing importance, rigorous research and evaluation are necessary (3, 4).
As healthcare is a distinctive domain intimately tied to the protection of human life and well-being, the utilization of AI should be considered. In addition, medical students constitute the backbone of the future healthcare workforce; therefore, examining their attitudes toward AI is highly important for guiding its responsible integration into health services and education. Recent evidence suggests that generative AI has already entered medical students’ learning practices. In a survey of German medical students, 76.2% reported using ChatGPT for medical education, most commonly for summarizing information, literature research, and clarifying concepts, highlighting its perceived efficiency while also raising concerns about misinformation and trust (5). Similarly, among graduating medical students surveyed in the United Arab Emirates, reported usage was lower in medical students, with anticipating use for exam preparation and exploring new medical topics (6). In the United States, recent work has begun to characterize frequency and motivations for ChatGPT use among medical students, reflecting growing integration into learning, writing, and research workflows. The most frequently used tools are LLM tools, typically applied for study support, drafting, and information synthesis rather than direct clinical decision-making (5). Importantly, policies aligned with its use exist: The Association of American Medical Colleges (AAMC) has issued principles to guide AI use in medical education, emphasizing transparency, privacy protection, and human-centered oversight (7), while universities increasingly publish guidance and syllabus statements for AI usage (8). At the governance level, United Nations Educational, Scientific and Cultural Organization (UNESCO) has released global guidance for AI in education and research (9). Taken together, these surveys and policy statements indicate that AI use is already embedded in medical training, while concerns about reliability, trust, and responsible use remain; understanding students’ attitudes and their determinants is therefore important for curriculum planning and governance.
Studies have investigated attitudes toward AI among both healthcare professionals and medical students. For healthcare professionals, an integrative review synthesizing 42 studies highlighted barriers such as concerns about autonomy, workflow fit, legal responsibility, and job displacement, alongside facilitators including training, usable interfaces, and organizational readiness (10). In addition, for medical students, a meta-analysis estimated that approximately 65% of medical, dental, and nursing students view AI positively, with marked between-country variability (11). This heterogeneity implies that beyond exposure and training, individual-level determinants may meaningfully shape how students evaluate AI. To support comparable measurement, the General Attitudes toward AI Scale (GAAIS) provides a robust two-factor (positive/negative) structure and has been validated across settings (12). Beyond educational exposure and contextual factors, stable individual differences may also shape medical students’ attitudes toward AI. In this study, we focus on personality traits as a theoretically grounded candidate determinant.
The classic Big Five personality framework classifies personality into five broad domains: openness, conscientiousness, extraversion, agreeableness and neuroticism (OCEAN). Openness reflects curiosity and a preference for novelty; conscientiousness captures self-discipline, organization, and reliability; extraversion indexes energetic social engagement; agreeableness denotes a prosocial, cooperative, and compassionate stance; neuroticism reflects a tendency toward negative affect, stress reactivity, and emotional lability. These dimensions, supported by decades of lexical and psychometric research, are stable and predictive of social behavior, well-being, and occupational and health outcomes (13). To assess these traits efficiently and with facet-level precision, the Big Five Inventory-2 (BFI-2) provides a 60-item measure that yields five domain scores and fifteen facet scores, enhancing bandwidth, fidelity, and interpretability in empirical research and practice (14). These traits may be relevant to AI adoption because they relate to how individuals approach novelty, uncertainty, and perceived risk, which may influence receptivity to AI and tolerance of its limitations.
An expanding body of literature explores how personality shapes attitudes toward AI in healthcare across cultures. In Turkey, dentistry students’ views on AI vary by personality profile (15), and work from Saudi Arabia suggests that all Big Five traits except extraversion were associated with scores of attitudes toward AI (16). In South America, studies indicate that personality dispositions influence healthcare workers’ willingness to adopt AI, with patterns conditioned by cultural norms and system structures (17). Research in African contexts underscores context-specific barriers to and facilitators of AI-enabled care, reinforcing the need for culturally responsive implementation strategies (18). Across Asia, rapid AI adoption in health services has prompted investigations among nursing students and professionals; these investigations link personality traits to AI acceptance and often diverge from Western findings (19–21). Nevertheless, to our knowledge, no study has yet examined the associations between personality and attitudes toward AI among Chinese medical students. Cross-cultural evidence suggests that attitudes toward AI are heterogeneous and may depend on sociocultural and educational contexts, making it inappropriate to assume that findings from other settings generalize directly to Chinese medical students (11). Although recent studies have begun to describe AI attitudes among medical students in China, the evidence remains limited and rarely examines psychological determinants such as personality traits using standardized attitude measures (22).
Our study aimed to investigate the associations between Chinese medical students’ attitudes toward AI and personality traits. The demographic data, personality traits and attitudes toward AI were collected from online questionnaires. The relationships between personality traits and attitudes toward AI were investigated through correlation analysis, and the personality traits that were associated with attitudes toward AI were investigated through linear regression. This study extends the emerging literature on psychological determinants of AI attitudes by examining whether Big Five traits are associated with attitudes toward AI in Chinese medical students, a context where evidence remains limited and cross-cultural patterns are heterogeneous. The findings also have practical value for medical education, as identifying trait-linked profiles of receptivity and caution can inform curriculum design and risk-aware training strategies for responsible AI use.
Materials and methods
Participants
A cross-sectional, web-based survey was conducted. Participants were recruited via convenience sampling from three public universities in Beijing, Shanghai, and Chengdu: Peking University Health Science Center; the School of Medicine, Shanghai Jiao Tong University; and the West China School of Clinical Medicine, Sichuan University from July to October, 2025. Eligible respondents were undergraduate medical students enrolled in a five-year clinical medicine program (the most common undergraduate track in China) at any year level who consented to complete the survey. The exclusion criteria were postgraduate status; current mental health difficulties or ongoing psychological treatment; and enrollment in self-improvement or professional development courses. To ensure that the target population completed the survey, eligibility was verified via self-report using mandatory screening items based on the inclusion and exclusion criteria.
The sample size was estimated with PASS (version 21.0.3, 2021) for a linear multiple regression framework. Using nine predictors (four demographic variables and the Big Five personality traits), an anticipated medium effect (f2 = 0.15) (23), α = 0.01, and 95% power, the minimum required N was 228. Considering an expected attrition rate of 12% for a web-based survey, the target sample size was inflated to 260 (24). Two attention check questions (If you could select “Strongly agree,” I would greatly appreciate it) were involved in the questionnaire (one in the GAAIS part and the other in the BFI-2 part). Participants selecting “Strongly agree” in both of the 2 attention check questions were included in our study. Ultimately, 253 medical students completed the web-based questionnaire after excluding 5 participants who did not pass the attention check, and were included in the final analysis, exceeding the minimum required sample size.
Demographic data
The demographic characteristics recorded included age, gender, marital status, and educational attainment.
GAAIS
The General Attitudes toward Artificial Intelligence Scale (GAAIS), developed by Schepman and Rodway (12), assesses overall dispositions toward AI. The original instrument contains 20 items across two subscales: positive attitudes (12 items) and negative attitudes (8 items). Subscales are scored separately, with higher values indicating stronger endorsement of positive or negative views. In the initial validation, internal consistency was high (α = 0.88 for the positive subscale; α = 0.83 for the negative subscale). To accommodate linguistic and cultural contexts, Huang et al. created a Chinese adaptation comprising 15 items (8 positive, 7 negative) (25). Positive items are rated on a five-point Likert scale (1 = strongly disagree to 5 = strongly agree); negative items are reverse-coded (1 = strongly agree to 5 = strongly disagree). Subscale scores are calculated as the means of their respective items. Higher scores on either subscale reflect more favorable attitudes toward AI; for the negative subscale, a higher mean denotes greater tolerance of AI limitations. The Chinese version used in our study showed good reliability, with Cronbach’s α ranging from 0.833–0.875, and convergent validity was supported by systematic correlations with measures of technology readiness and personality (25).
BFI-2
The Big Five Inventory (BFI), developed by Benet-Martínez and John, assesses five broad personality domains with 44 items: extraversion (8), agreeableness (9), conscientiousness (9), neuroticism (8), and openness (10) (26). The revised BFI-2 expands the instrument to 60 items, assigning 12 items to each domain (14). Zhang et al. translated BFI-2 into Chinese, which was applied in our study, and demonstrated its good reliability, structural validity, convergent/discriminant validity, and criterion-related validity at the domain level (27).
Data collection and ethics
Data were gathered via a questionnaire hosted on WenJuanXing, a widely used online survey platform. The link was disseminated mainly via instant messaging tools in China, such as WeChat and QQ. The opening page provided an invitation describing the study objectives and intended participant group. Participation was voluntary, with assurances of confidentiality. Electronic informed consent was obtained; respondents had to click “I agree” before proceeding. To maintain data quality, all the items were compulsory, and the system prevented the submission of incomplete forms. The survey was anonymous and emphasized that there were no right or wrong answers to reduce social desirability and evaluation apprehension. The study protocol was approved by the Ethics Committee of West China Hospital, Sichuan University (Approval No. 2024–0371).
Statistical analyses
All analyses were conducted in SPSS 25.0. Participant demographics and study variables were summarized using frequencies and percentages, as well as means and standard deviations. Procedures were implemented to evaluate the measures’ validity and reliability. The common method variance (CMV) was assessed using Harman’s single-factor test (unrotated principal component factor analysis of all GAAIS and BFI-2 items). Between-group differences by demographic factors were tested with independent-samples t tests and one-way ANOVA. Associations between personality dimensions and attitudes toward AI were examined with Pearson correlation coefficients. Finally, stepwise multiple linear regressions were fitted separately for the positive and negative attitude scores.
Results
Participant demographics
The participants were approximately evenly distributed by age, gender, and year of study; all the participants were unmarried, reflecting the rarity of marriage among Chinese undergraduates. No statistically significant demographic differences were observed in medical students’ attitudes toward AI (Table 1).
Table 1.
Participant demographics and between-group differences in attitudes toward AI (N = 253).
| Category | N | % | Positive | Negative | |||
|---|---|---|---|---|---|---|---|
| Mean (SD) | t/F (p value) | Mean (SD) | t/F (p value) | ||||
| Age | <20 | 106 | 41.9 | 2.62 (1.01) | t = 1.50 (0.14) | 3.15 (1.06) | t = −1.40 (0.162) |
| ≥20 | 147 | 58.1 | 2.42 (1.07) | 3.36 (1.20) | |||
| Gender | Male | 116 | 45.8 | 2.63 (1.01) | t = 1.84 (0.066) | 3.41 (1.18) | t = 1.73 (0.085) |
| Female | 137 | 54.2 | 2.39 (1.07) | 3.16 (1.11) | |||
| Marital status | Married | 0 | 0.0 | NA | NA | NA | NA |
| Unmarried | 253 | 100.0 | 2.50 (1.05) | 3.27 (1.15) | |||
| Educational level | 1st year | 39 | 15.4 | 2.39 (0.94) | F = 1.87 (0.117) | 3.13 (1.09) | F = 1.087 (0.363) |
| 2nd year | 68 | 26.9 | 2.70 (1.05) | 3.24 (1.02) | |||
| 3rd year | 54 | 21.3 | 2.22 (1.09) | 3.48 (1.24) | |||
| 4th year | 59 | 23.3 | 2.61 (1.03) | 3.36 (1.18) | |||
| 5th year | 33 | 13.0 | 2.50 (1.07) | 3.02 (1.24) | |||
Descriptive statistics and reliability
Table 2 shows that negative attitudes (M = 3.27, SD = 1.15) scored more than positive attitudes did (M = 2.50, SD = 1.05). The mean scores for personality traits were as follows: openness (M = 3.53, SD = 0.62), conscientiousness (M = 3.21, SD = 0.44), extraversion (M = 3.18, SD = 0.43), agreeableness (M = 3.46, SD = 0.63), and neuroticism (M = 2.90, SD = 0.39). Harman’s single-factor test indicated that the first factor accounted for 11.36% of the total variance, suggesting that CMV was unlikely to substantially bias the observed associations.
Table 2.
Descriptive statistics and reliability (N = 253).
| Variable | Mean (SD) | Cronbach’s α |
|---|---|---|
| Attitude toward AI | ||
| Positive attitude | 2.50 (1.05) | 0.85 |
| Negative attitude | 3.27 (1.15) | 0.88 |
| Personality traits | ||
| Openness | 3.53 (0.62) | 0.84 |
| Conscientiousness | 3.21 (0.44) | 0.89 |
| Extraversion | 3.18 (0.43) | 0.76 |
| Agreeableness | 3.46 (0.63) | 0.85 |
| Neuroticism | 2.90 (0.39) | 0.76 |
Reliability was evaluated via Cronbach’s α, and the α coefficients ranged from 0.76–0.89 across all factors, exceeding the conventional 0.70 benchmark and indicating strong internal consistency.
Correlation analysis between attitudes toward AI and personality traits
Table 3 reports the Pearson correlation results. Positive attitudes toward AI were significantly associated with greater openness (r = 0.37, p < 0.001) and agreeableness (r = 0.19, p = 0.003) and with lower conscientiousness (r = −0.21, p = 0.001). For negative attitudes, openness (r = 0.25, p < 0.001) and agreeableness (r = 0.33, p < 0.001) had positive associations, whereas neuroticism had a negative association (r = −0.26, p < 0.001). The correlation coefficient and p value were visualized in Figure 1.
Table 3.
The correlations between attitudes toward AI and personality traits (N = 253).
| Personality traits | Attitudes toward AI | |
|---|---|---|
| Positive | Negative | |
| Openness | ||
| Correlation coefficient r | 0.37 | 0.25 |
| p value | <0.001 | <0.001 |
| Conscientiousness | ||
| Correlation coefficient r | −0.21 | 0.004 |
| p value | 0.001 | 0.954 |
| Extraversion | ||
| correlation coefficient r | −0.031 | 0.013 |
| p value | 0.628 | 0.835 |
| Agreeableness | ||
| correlation coefficient r | 0.19 | 0.33 |
| p value | 0.003 | <0.001 |
| Neuroticism | ||
| correlation coefficient r | −0.038 | −0.26 |
| p value | 0.552 | <0.001 |
Figure 1.
The heatmap showed the correlation between attitudes toward AI and personality traits. Redder color indicates stronger positive correlations while bluer color indicates stronger negative correlations. *p < 0.05, **p < 0.01, ***p < 0.001.
Linear regression analysis for predictors of attitudes toward AI
Hierarchical multiple linear regression predicting attitudes toward AI was conducted: demographic covariates (age, gender, and educational level dummies) were forced into Block 1 using the Enter method to control potential confounding, and the personality traits that were significant in the correlation analyses were selected in Block 2 using stepwise procedures.
In hierarchical regression predicting positive attitudes toward AI, Block 1 (age, gender, and educational level) was not significant (adjusted R2 = 0.002, p = 0.32). In Block 2, stepwise selection retained openness, agreeableness, and conscientiousness as significant predictors. Adding openness significantly improved model fit (ΔR2 = 0.1, p < 0.001), followed by agreeableness (ΔR2 = 0.04, p = 0.001) and conscientiousness (ΔR2 = 0.084, p < 0.001). The final model explained 24.7% of the variance in positive attitudes (adjusted R2 = 0.247, p < 0.001). In the final model, openness (β = 0.32, p < 0.001) and agreeableness (β = 0.38, p < 0.001) were positively associated with positive attitudes, whereas conscientiousness was negatively associated (β = −0.35, p < 0.001) (Table 4; Figure 2). Age, gender, and educational level were not significant. Collinearity diagnostics were conducted, the tolerance values ranged from 0.70–0.95, with VIFs ranging from 1.05–1.43, well above/below the conventional 0.10- and 3.0 cutoffs, respectively, and no multicollinearity issues were found for the positive-attitude models. Model-building details for Models 1–3 are provided in Supplementary Table S1.
Table 4.
Final model for stepwise multiple linear regression results identifying predictors of medical students’ positive attitudes toward AI.
| Predictors | B (95% CI) | β | t | p value |
|---|---|---|---|---|
| Age | 0.021 (−0.25, 0.29) | 0.035 | 0.15 | 0.88 |
| Gender | 0.068 (−0.16, 0.30) | 0.032 | 0.58 | 0.56 |
| Educational level | −0.035 (−0.41, 0.34) | −0.043 | −0.19 | 0.85 |
| Openness | 0.54 (0.36, 0.73) | 0.32 | 5.71 | <0.001 |
| Conscientiousness | −0.82 (−1.13, −0.52) | −0.35 | −5.86 | <0.001 |
| Agreeableness | 0.63 (0.42, 0.84) | 0.38 | 5.30 | <0.001 |
B, unstandardized regression coefficient; CI, confidence interval; t, t-statistics.
Figure 2.
The forest plot showed the regression coefficient B of personality traits in the regression model for positive attitude.
In hierarchical regression predicting negative attitudes toward AI, Block 1 (age, gender, and educational level) showed a small amount of explained variance and did not reach statistical significance (adjusted R2 = 0.016, p = 0.074). In Block 2, stepwise selection retained agreeableness, neuroticism, and openness as significant predictors. Adding agreeableness significantly improved model fit (ΔR2 = 0.110, p < 0.001), followed by neuroticism (ΔR2 = 0.077, p < 0.001) and openness (ΔR2 = 0.059, p < 0.001). The final model explained 25.7% of the variance in negative attitudes (adjusted R2 = 0.257, p < 0.001). In the final model, agreeableness (β = 0.36, p < 0.001) and openness (β = 0.25, p < 0.001) were positively associated with negative attitudes, whereas neuroticism was negatively associated (β = −0.28, p < 0.001). Educational level showed a small negative association (β = −0.46, p = 0.048), while age and gender were not significant in the final model (Table 5; Figure 3). Collinearity diagnostics were conducted, the tolerance values ranged from 0.98–0.99, with VIFs ranging from 1.015–1.024, well above/below the conventional 0.10- and 3.0 cutoffs, respectively, and no multicollinearity issues were found for the negative-attitude models. Detailed coefficients for Models 1–3 are provided in Supplementary Table S2.
Table 5.
Stepwise multiple linear regression results identifying predictors of medical students’ negative attitudes toward AI.
| Predictors | B (95% CI) | β | t | p value |
|---|---|---|---|---|
| Age | 0.27 (−0.029, 0.57) | 0.41 | 1.78 | 0.077 |
| Gender | 0.17 (−0.077, 0.42) | 0.075 | 1.36 | 0.18 |
| Educational level | −0.41 (−0.82, −0.003) | −0.46 | −1.98 | 0.048 |
| Openness | 0.46 (0.26, 0.66) | 0.25 | 4.49 | <0.001 |
| Agreeableness | 0.66 (0.47, 0.86) | 0.36 | 6.63 | <0.001 |
| Neuroticism | −0.83 (−1.15, −0.51) | −0.28 | −5.08 | <0.001 |
B, unstandardized regression coefficient; CI, confidence interval; t, t-statistics.
Figure 3.
The forest plot showed the regression coefficient B of personality traits in the regression model for negative attitude.
Discussion
This study investigated how personality traits shape Chinese medical students’ attitudes toward AI. Although prior work has assessed healthcare professionals’ views, evidence focused on medical students in China is sparse. Drawing on the BFI framework, we examine whether specific traits were associated with variation in attitudes toward AI, addressing this gap. Overall, attitudes toward AI among Chinese medical students appeared less differentiated by demographics and more closely related to personality. The findings suggested that openness and agreeableness were linked to greater receptivity toward AI, while conscientiousness and neuroticism were linked to more reserved evaluations, highlighting the value of tailoring AI education to individual differences. Our analyses revealed that demographic characteristics appeared to play a limited role in shaping attitudes toward AI. This aligns with prior evidence that the cultural context shapes technology perceptions and that a shared cultural background may yield similar views. This interpretation is consistent with prior work suggesting that cultural context and shared norms can strongly influence how new technologies are understood and evaluated (17, 28, 29). In contrast, individual psychological differences, particularly personality traits, may become more salient in explaining why some students are more receptive to AI whereas others remain more cautious.
Consistent with prior research, individuals high in openness are typically curious, innovation oriented, and receptive to new technologies. This disposition may make students more willing to experiment with AI, view it as a tool for learning and improvement, and remain more comfortable with its inevitable imperfections in clinical use (30, 31). Similarly, Salem et al. reported that openness was associated with higher scores of positive attitudes toward AI among nursing students (16). Taken together, this pattern supports the broader view that openness facilitates constructive engagement with emerging technologies during healthcare training.
Moreover, agreeableness may be linked to more favorableness toward AI and greater tolerance of its shortcomings. Because agreeableness reflects compassion and cooperativeness, students who score higher on this trait may adopt AI more readily and show greater tolerance of its limitations than their less agreeable peers do. Nonetheless, this pattern differs from that of Salem et al. and other studies reporting no significant link between agreeableness and positive attitudes toward AI. The cultural context may help explain this discrepancy. East Asian students, shaped by collectivist norms from early childhood, tend to value cooperation, which may promote greater tolerance (32).
Conscientiousness may be associated with more reserved evaluations of AI among medical students, particularly when students consider the risks and uncertainties of clinical use. This pattern may reflect greater meticulousness and adherence to protocols, which in turn fosters heightened cautiousness of AI limitations. However, evidence elsewhere is mixed. Kaya et al. reported no predictive effect of conscientiousness on either positive or negative attitudes (30), whereas other studies reported a weak but significant positive association with negative attitudes (33, 34). These discrepancies underscore the need for further investigation.
Furthermore, neuroticism may reduce students’ tolerance for AI’s shortcomings by amplifying anxiety, skepticism, and perceived threat in response to AI-related uncertainty. This finding is consistent with prior evidence that highly neurotic individuals are more prone to anxiety, skepticism, and fear regarding AI risks and implications (31). They may perceive AI as a threat rather than a tool, and the prospect of machines emulating human capabilities and displacing human tasks can amplify the insecurity and job-loss concerns of these participants, thereby reinforcing unfavorable attitudes (35, 36).
However, extraversion appeared to play a limited role in shaping medical students’ attitudes toward AI, which is consistent with prior reports showing no significant associations (16, 30, 31, 33). One possible explanation is that extraversion may be less directly relevant to evaluations of AI as a clinical tool, with its influence depending on more interpersonal or social-use contexts. In addition, within an East Asian collectivist cultural background, where social harmony and solidarity are emphasized over individual expressiveness, differences related to extraversion may be less salient in shaping attitudes toward AI.
This study has several limitations. First, the sample was restricted to students in five-year clinical medicine programs, the most common undergraduate track in China, so the findings may not be generalizable to 7- or 8-year programs or to dental students, and future work should include these groups. Second, the cross-sectional design limits causal inference, and longitudinal studies are needed to track how personality relates to attitudes over time and across training and clinical contexts. Third, all variables were collected via self-report, which may be subject to recall and social desirability bias. CMV is a potential concern given the self-report design. Nevertheless, Harman’s single-factor test provided some reassurance (the first factor explained 11.36% of total variance), although CMV cannot be fully excluded. In addition, although we used Chinese versions with published psychometric support, we did not test cross-cultural measurement invariance of the GAAIS or BFI-2; therefore, the present findings should not be directly generalized to other cultural contexts or used for cross-country comparisons without further equivalence testing. Furthermore, this study used a web-based convenience sample recruited via social media, which may introduce selection and coverage bias. Students who were more interested in AI or more engaged online may have been more likely to participate, limiting representativeness and generalizability. Although we recruited participants from three universities in different regions, the sample may not reflect all Chinese medical students. Future studies should consider probability-based or stratified sampling and mixed-mode (online/offline) recruitment.
Implications
To our knowledge, this is the first study to examine the associations between personality traits and attitudes toward AI among Chinese medical students. Taken together, the findings suggest that individual differences may shape whether students approach AI with curiosity and openness to collaboration, or with greater caution when considering uncertainty and potential risks (37). From a curricular perspective, AI content is likely to be most effective when it is integrated longitudinally across the existing program (introduced early and revisited in clinically anchored contexts), rather than delivered as a single, stand-alone module (38). This integration can be operationalized through specialty-relevant cases and structured critical-appraisal activities, and consolidated using simulation or OSCE-style tasks that require safe human-AI collaboration, documentation of AI-assisted reasoning, and communication with patients about uncertainty, ethics, and governance. Pedagogically, tailoring can be modest but practical: students who are more receptive to AI may benefit from team-based projects (e.g., workflow mapping, structured evaluation of a published clinical AI tool, or guided use-case development), whereas students who are more cautious or less tolerant of AI shortcomings may benefit from stepwise exposure, verification checklists, error-analysis exercises, and supported debriefing to build calibrated trust without overreliance (37). In particular, for groups with lower tolerance, teaching should explicitly emphasize competencies repeatedly highlighted in recent competency frameworks and curriculum guidance-recognizing limitations and failure modes, uncertainty calibration and verification, bias/fairness awareness, ethical and legal accountability, and monitoring AI use in real clinical workflows (39).
Conclusion
In our study, the data on attitudes toward AI were collected through the GAAIS, and personality trait data were collected through the BFI-2, both via an online system. The correlations between attitudes toward AI and personality traits were explored. The traits that were significantly correlated with positive and negative attitudes were subsequently selected via linear regression. In the regression model, we found that for positive attitudes toward AI, openness and agreeableness were associated with its higher scores, whereas conscientiousness was associated with its lower scores. For negative attitudes toward AI, openness and agreeableness were associated with its higher scores, whereas neuroticism was associated with its lower scores. These findings suggested that students with high openness and agreeableness not only had more positive attitudes toward AI but were also more tolerant of the shortcomings of AI. Students with high conscientiousness had fewer positive attitudes toward AI, and students with high neuroticism were not tolerant of the shortcomings of AI. Our study is the first to investigate the relationship between attitudes toward AI and personality traits in medical students, suggesting the need for targeted educational interventions that consider the diverse personality traits of medical students.
Acknowledgments
We thank all the members of Laboratory of Dermatology who assisted in this course and valuable suggestions on this manuscript.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the National Natural Science Foundation of China (82404175), the Funds of Science & Technology Department of Sichuan Province (2025ZNSFSC1629 and 24NSFSC1625), and the Research Fund of West China Hospital, Sichuan University (2023HXBH032 and 2025HXBH117).
Footnotes
Edited by: Dirk Bruland, Bielefeld University of Applied Sciences, Germany
Reviewed by: Ahmad Samed Al-Adwan, Al-Ahliyya Amman University, Jordan
Glenn Ford Valdez, Oman College of Health Science -Dhofar, Oman
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving humans were approved by Ethics Committee of West China Hospital, Sichuan University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
JQ: Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. YZe: Data curation, Formal analysis, Investigation, Methodology, Software, Writing – original draft, Writing – review & editing. HW: Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. YX: Investigation, Methodology, Writing – original draft, Writing – review & editing. ZF: Data curation, Investigation, Writing – original draft, Writing – review & editing. YZh: Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. TB: Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. SY: Data curation, Investigation, Methodology, Writing – original draft, Writing – review & editing. LL: Data curation, Investigation, Supervision, Writing – original draft, Writing – review & editing. YS: Conceptualization, Data curation, Investigation, Methodology, Writing – original draft, Writing – review & editing. XJ: Writing – review & editing, Supervision, Conceptualization. SC: Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Supervision, Writing – original draft, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh.2026.1749279/full#supplementary-material
References
- 1.Bohr A, Memarzadeh K. The rise of artificial intelligence in healthcare applications In: Bohr A, Memarzadeh K, editors. Artificial intelligence in healthcare. London: Academic Press; (2020). 25–60. [Google Scholar]
- 2.Ali O, Abdelbaki W, Shrestha A, Elbasi E, Alryalat MAA, Dwivedi YK. A systematic literature review of artificial intelligence in the healthcare sector: benefits, challenges, methodologies, and functionalities. J Innov Knowl. (2023) 8:100333. doi: 10.1016/j.jik.2023.100333 [DOI] [Google Scholar]
- 3.Adams L, Fontaine E, Lin S, Crowell T, Chung VCH, Gonzalez AA. Artificial intelligence in health, health care, and biomedical science: an AI code of conduct principles and commitments discussion draft. NAM Perspect. (2024) 2024:202403a. doi: 10.31478/202403a [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Goodman KW, Litewka SG, Malpani R, Pujari S, Reis AA. Global health and big data: the WHO'S artificial intelligence guidance. S Afr J Sci. (2023) 119:14725. doi: 10.17159/sajs.2023/14725 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fußhöller A, Lechner F, Schlicker N, Muehlensiepen F, Mayr A, Kuhn S, et al. Perceptions, usage, and educational impact of ChatGPT among medical students in Germany: cross-sectional mixed methods survey. JMIR Form Res. (2025) 9:e81484. doi: 10.2196/81484, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Alkhaaldi SMI, Kassab CH, Dimassi Z, Oyoun Alsoud L, Al Fahim M, Al Hageh C, et al. Medical student experiences and perceptions of ChatGPT and artificial intelligence: cross-sectional study. JMIR Med Educ. (2023) 9:e51302. doi: 10.2196/51302, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Association of American Medical Colleges Principles for the responsible use of artificial intelligence in and for medical education (2025). Available online at: https://www.aamc.org/about-us/mission-areas/medical-education/principles-ai-use (Accessed December 20, 2025).
- 8.Johns Hopkins University Generative AI tool guidance (2025). Available online at: https://teaching.jhu.edu/university-teaching-policies/generative-ai/ (Accessed December 20, 2025).
- 9.Miao F, Holmes W. Guidance for generative AI in education and research [Internet]. Paris: UNESCO; (2023). Available online at: https://unesdoc.unesco.org/ark:/48223/pf0000386693 (Accessed December 20, 2025). [Google Scholar]
- 10.Lambert SI, Madi M, Sopka S, Lenes A, Stange H, Buszello CP, et al. An integrative review on the acceptance of artificial intelligence among healthcare professionals in hospitals. NPJ Digit Med. (2023) 6:111. doi: 10.1038/s41746-023-00852-5, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Amiri H, Peiravi S, Rezazadeh Shojaee SS, Rouhparvarzamin M, Nateghi MN, Etemadi MH, et al. Medical, dental, and nursing students' attitudes and knowledge towards artificial intelligence: a systematic review and meta-analysis. BMC Med Educ. (2024) 24:412. doi: 10.1186/s12909-024-05406-1, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Schepman A, Rodway P. Initial validation of the general attitudes towards artificial intelligence scale. Comput Hum Behav Rep. (2020) 1:100014. doi: 10.1016/j.chbr.2020.100014, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zell E, Lesick TL. Big five personality traits and performance: a quantitative synthesis of 50+ meta-analyses. J Pers. (2022) 90:559–73. doi: 10.1111/jopy.12683, [DOI] [PubMed] [Google Scholar]
- 14.Soto CJ, John OP. The next big five inventory (BFI-2): developing and assessing a hierarchical model with 15 facets to enhance bandwidth, fidelity, and predictive power. J Pers Soc Psychol. (2017) 113:117–43. doi: 10.1037/pspp0000096, [DOI] [PubMed] [Google Scholar]
- 15.Ozbey F, Yasa Y. The relationships of personality traits on perceptions and attitudes of dentistry students towards AI. BMC Med Educ. (2025) 25:26. doi: 10.1186/s12909-024-06630-5, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Salem GMM, El-Gazar HE, Mahdy AY, Alharbi TAF, Zoromba MA. Nursing students' personality traits and their attitude toward artificial intelligence: a multicenter cross-sectional study. J Nurs Manag. (2024) 2024:6992824. doi: 10.1155/2024/6992824 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Riedl R. Is trust in artificial intelligence systems related to user personality? Review of empirical evidence and future research directions. Electron Mark. (2022) 32:2021–51. doi: 10.1007/s12525-022-00594-4 [DOI] [Google Scholar]
- 18.Ohalete NC, Ayo-Farai O, Olorunsogo TO, Maduka P, Olorunsogo T. AI-driven environmental health disease modeling: a review of techniques and their impact on public health in the USA and African contexts. Int Med Sci Res J. (2024) 4:51–73. doi: 10.51594/imsrj.v4i1.737 [DOI] [Google Scholar]
- 19.Labrague LJ, Aguilar-Rosales R, Yboa BC, Sabio JB. Factors influencing student nurses' readiness to adopt artificial intelligence (AI) in their studies and their perceived barriers to accessing AI technology: a cross-sectional study. Nurse Educ Today. (2023) 130:105945. doi: 10.1016/j.nedt.2023.105945, [DOI] [PubMed] [Google Scholar]
- 20.Hamedani Z, Moradi M, Kalroozi F, Manafi Anari A, Jalalifar E, Ansari A, et al. Evaluation of acceptance, attitude, and knowledge towards artificial intelligence and its application from the point of view of physicians and nurses: a provincial survey study in Iran: a cross-sectional descriptive-analytical study. Health Sci Rep. (2023) 6:e1543. doi: 10.1002/hsr2.1543, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Atta MHR, Zoromba MA, El-Gazar HE, Loutfy A, Elsheikh MA, El-Ayari OSM, et al. Climate anxiety, environmental attitude, and job engagement among nursing university colleagues: a multicenter descriptive study. BMC Nurs. (2024) 23:133. doi: 10.1186/s12912-024-01788-1, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Liu M, Cheng Y, Li S, Wang S, Du F, Wei X, et al. Exploring medical students' attitudes and perceptions toward artificial intelligence in medicine in Shandong Province, China. BMC Med Educ. (2025). doi: 10.1186/s12909-025-08465-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Groß J, Möller A. Effect size estimation in linear mixed models. Metron. (2025) 83:353–63. doi: 10.1007/s40300-025-00295-w [DOI] [Google Scholar]
- 24.Hoerger M. Participant dropout as a function of survey length in internet-mediated university studies: implications for study design and voluntary participation in psychological research. Cyberpsychol Behav Soc Netw. (2010) 13:697–700. doi: 10.1089/cyber.2009.0445, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Huang Y, Jiang S, Gong Z. Validity and reliability of the Chinese version of general attitudes towards artificial intelligence scale. Int J Hum Comput Interact. (2025) 41:12884–94. doi: 10.1080/10447318.2025.2465868 [DOI] [Google Scholar]
- 26.Benet-Martínez V, John OP. Los Cinco Grandes across cultures and ethnic groups: multitrait multimethod analyses of the big five in Spanish and English. J Pers Soc Psychol. (1998) 75:729–50. doi: 10.1037//0022-3514.75.3.729 [DOI] [PubMed] [Google Scholar]
- 27.Zhang B, Li YM, Li J, Luo J, Ye Y, Yin L, et al. The big five Inventory-2 in China: a comprehensive psychometric evaluation in four diverse samples. Assessment. (2022) 29:1262–84. doi: 10.1177/10731911211008245, [DOI] [PubMed] [Google Scholar]
- 28.Buchanan C, Howitt ML, Wilson R, Booth RG, Risling T, Bamford M. Predicted influences of artificial intelligence on the domains of nursing: scoping review. JMIR Nurs. (2020) 3:e23939. doi: 10.2196/23939, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Mousavi Baigi SF, Sarbaz M, Ghaddaripouri K, Ghaddaripouri M, Mousavi AS, Kimiafar K. Attitudes, knowledge, and skills towards artificial intelligence among healthcare students: a systematic review. Health Sci Rep. (2023) 6:e1138. doi: 10.1002/hsr2.1138, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kaya F, Aydin F, Schepman A, Rodway P, Yetişensoy O, Demir Kaya M. The roles of personality traits, AI anxiety, and demographic factors in attitudes toward artificial intelligence. Int J Hum Comput Interact. (2024) 40:497–514. doi: 10.1080/10447318.2022.2151730 [DOI] [Google Scholar]
- 31.Park J, Woo SE. Who likes artificial intelligence? Personality predictors of attitudes toward artificial intelligence. J Psychol. (2022) 156:68–94. doi: 10.1080/00223980.2021.2012109, [DOI] [PubMed] [Google Scholar]
- 32.Rudolf R. Predictors of adolescent well-being around the globe: are students from Confucian East Asia different? Front Psychol. (2024) 15:1446301. doi: 10.3389/fpsyg.2024.1446301, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Sindermann C, Riedl R, Montag C. Investigating the relationship between personality and technology acceptance with a focus on the smartphone from a gender perspective: results of an exploratory survey study. Future Internet. (2020) 12:110. doi: 10.3390/fi12070110 [DOI] [Google Scholar]
- 34.Kwak Y, Ahn J-W, Seo YH. Influence of AI ethics awareness, attitude, anxiety, and self-efficacy on nursing students' behavioral intentions. BMC Nurs. (2022) 21:267. doi: 10.1186/s12912-022-01048-0, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Cho KA-O, Seo YA-O. Dual mediating effects of anxiety to use and acceptance attitude of artificial intelligence technology on the relationship between nursing students' perception of and intention to use them: a descriptive study. BMC Nurs. (2024) 23:212. doi: 10.1186/s12912-024-01887-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Sindermann C, Yang H, Elhai JD, Yang S, Quan L, Li M, et al. Acceptance and fear of artificial intelligence: associations with personality in a German and a Chinese sample. Discov Psychol. (2022) 2:8. doi: 10.1007/s44202-022-00020-y [DOI] [Google Scholar]
- 37.Stein JP, Messingschlager T, Gnambs T, Hutmacher F, Appel M. Attitudes towards AI: measurement and associations with personality. Sci Rep. (2024) 14:2909. doi: 10.1038/s41598-024-53335-2, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Kim DH, Kang YJ, Lee YM. Twelve tips for developing and implementing AI curriculum for undergraduate medical education. Med Educ Online. (2025) 30:2585637. doi: 10.1080/10872981.2025.2585637, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Car J, Ong QC, Erlikh Fox T, Leightley D, Kemp SJ, Švab I, et al. The digital health competencies in medical education framework: an international consensus statement based on a delphi study. JAMA Netw Open. (2025) 8:e2453131. doi: 10.1001/jamanetworkopen.2024.53131, [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.