Generative AI Chatbots in Chinese Postgraduate Medical Education: Adoption Patterns, Attitudes, and Concerns

Generative AI Chatbots in Chinese Postgraduate Medical Education: Adoption Patterns, Attitudes, and Concerns | 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 Generative AI Chatbots in Chinese Postgraduate Medical Education: Adoption Patterns, Attitudes, and Concerns Lijun Zhao, Qianqian Han, Peijuan Li, Hua Dai, Xuegui Ju This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7461983/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 18 Mar, 2026 Read the published version in BMC Medical Education → Version 1 posted 13 You are reading this latest preprint version Abstract Background Generative artificial intelligence (AI) chatbots are gaining attention in medical education for their potential to support academic writing, clinical reasoning, and personalized learning. However, little is known about their adoption, benefits, and concerns among Chinese postgraduate medical students, particularly across distinct clinical- and academic-track programs. Methods A cross-sectional survey was conducted among 340 postgraduate medical students (146 males, 194 females) from two universities in Chengdu, China. A structured questionnaire assessed AI awareness, usage patterns, perceived benefits, attitudes, and concerns. Descriptive statistics, sensitivity analyses, and Pearson correlation analyses were applied to examine differences by gender and degree type. Results Most students (82.9%) reported strong AI awareness, with DeepSeek (90.9%) and ChatGPT (55.2%) being the most frequently used tools. Common applications included literature review (61.5%), exam preparation (55.0%), clinical case analysis (48.5%), and academic writing (43.2%). Reported benefits encompassed faster information retrieval (70.8%), improved writing precision (64.3%), and greater confidence in clinical decision-making (58.0%). Overall satisfaction was high (mean 4.4/5), and 84.0% of participants supported integration of AI into medical curricula. Female students reported more frequent use but greater concerns, while clinical-track students demonstrated higher awareness and satisfaction than academic-track students. Correlation analyses revealed positive associations among awareness, usage, perceptions, and attitudes, whereas concerns about accuracy, academic misconduct, and data security were largely independent of prior exposure. Conclusions Generative AI chatbots are widely adopted and valued by Chinese postgraduate medical students. Structured integration into medical curricula, accompanied by clear ethical safeguards, is essential to maximize benefits while addressing potential risks. Generative artificial intelligence Chatbots Medical education Technology adoption Perceptions Postgraduate medical students China Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Artificial Intelligence (AI) refers to computational systems capable of performing tasks that typically require human intelligence, encompassing domains such as machine learning (ML), natural language processing (NLP), and deep learning. In medicine, AI has progressed from theoretical models to clinically integrated tools, transforming fields such as medical imaging analysis (e.g., tumor detection in radiology), predictive diagnostics (e.g., sepsis risk stratification), drug discovery, and personalized treatment planning[ 1 , 2 ]. The World Health Organization has acknowledged AI’s potential to address global healthcare disparities, particularly in resource-limited settings[ 3 ]. Generative AI—an emerging subset of AI capable of producing novel text, images, or structured data—represents a significant frontier with transformative implications for both clinical practice and medical education. Medical chatbots, powered by NLP, have traditionally been used to simulate human conversation, assist in scheduling, and manage electronic health records. More recently, generative AI chatbots such as ChatGPT and DeepSeek have demonstrated the ability to synthesize complex, context-aware responses to open-ended academic and clinical queries, thereby extending their applications beyond routine administrative tasks[ 4 ]. AI-based educational tools, including intelligent virtual patient simulations, adaptive learning platforms, and generative question–answer systems, are increasingly being integrated into medical curricula, with evidence suggesting improvements in teaching efficiency and student engagement[ 5 ]. The adoption of generative AI chatbots in medical education offers several advantages, including enhanced access to information, real-time language refinement, and personalized feedback[ 6 ]. However, legitimate concerns persist, particularly regarding the generation of inaccurate information (“hallucinations”), the potential for plagiarism, biases in training data, and copyright violations[ 7 , 8 ]. Such issues may influence the perceptions, expectations, and attitudes of both students and educators, potentially shaping the effectiveness of AI-enhanced teaching[ 9 ]. Previous research has suggested that postgraduate medical students may exhibit higher awareness of AI tools compared to undergraduates[ 9 ]. Nevertheless, systematic and comparative analyses of AI awareness and adoption across different phases of medical education remain limited. Moreover, although international studies have explored AI adoption in medical training[ 10 ], China’s medical education system—distinguished by clear differences between clinical and academic degree pathways—remains underexamined. Clinical postgraduate programs (M.D./M.Med.) emphasize hands-on hospital training across specialties, enabling graduates to transition directly into clinical practice with residency certification. In contrast, academic postgraduate programs (Ph.D./M.Sc.) focus on laboratory research, scientific publishing, and theoretical study, preparing graduates for careers in research or academia, but requiring additional clinical training for hospital-based roles. The fundamental distinction lies in training emphasis—patient care versus research—and the readiness for specific career pathways after graduation. How these differing forms of clinical exposure shape students’ adoption of generative AI chatbots is not yet well understood. To address this gap, the present cross-sectional study investigates adoption patterns of generative chatbots among Chinese medical students, evaluates their attitudes and concerns, and explores behavioral intentions toward AI in both academic and clinical learning contexts, thereby providing evidence tailored to China’s unique medical education landscape. Materials and methods Study design This cross-sectional survey was conducted among postgraduate medical students from two universities in Chengdu, China, with the aim of assessing their knowledge, attitudes, concerns, and behavioral intentions toward the use of generative AI chatbots in academic and educational contexts. Data were collected between June and August 2025. The questionnaire was distributed to students across different specialties and stages of postgraduate training. Eligible participants were current postgraduate medical students who had access to electronic devices and consented to participate in the study. Exclusion criteria were applied to individuals who were not enrolled in medical universities, lacked internet access, or declined participation. This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board at the West China Hospital of Sichuan University. All participants provided informed consent prior to their inclusion in the study. Electronic consent was obtained from every participant in accordance with institutional and ethical guidelines. Questionnaire content The online survey was developed in accordance with the Checklist for Reporting Results of Internet E-Surveys (CHERRIES) [ 11 ] and followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines[ 12 ]. The final questionnaire comprised 40 items organized into seven sections and was administered via the WJX platform ( https://www.wjx.cn/ ). The survey addressed the following domains: participant demographics, awareness and understanding of AI, patterns of AI chatbot use, educational applications of AI, perceptions of generative AI, and students’ attitudes, concerns, and expectations regarding AI in medical education and academic research. The survey link was disseminated through community engagement groups on the WeChat social media platform (Tencent Inc., Shenzhen, China). Data cleaning and quality control To ensure data integrity and reliability, responses underwent a structured cleaning process. Exclusion criteria were applied as follows: (1) responses with excessive missing data, defined as more than one unanswered item; (2) submissions demonstrating uniform or highly repetitive answer patterns suggestive of inattentive responding; and (3) completion times shorter than 50 seconds for the entire questionnaire or under 30 seconds after answering the first 11 questions. These thresholds were established through a pilot test involving five medical students from the research team, who determined benchmarks for reasonable completion times. A total of 345 students participated, of which 5 responses met the exclusion criteria and were removed from the final dataset. Statistical Analysis Survey data were exported from the WJX platform into Microsoft Excel for initial management and subsequently analyzed using Stata software (version 17.0, StataCorp, College Station, TX, USA). Descriptive statistics were used to summarize the data: categorical variables were presented as frequencies and percentages, and continuous variables as means with standard deviations (SD). Independent samples t -tests were applied to compare continuous variables between groups, while chi-square (χ²) tests were used for categorical variables. Pearson correlation analysis was conducted to explore relationships among students’ awareness, frequency of AI usage, perceptions, attitudes, and concerns regarding AI in medical education. A two-tailed p-value of < 0.05 was considered indicative of statistical significance. Results Participant demographics A total of 340 medical postgraduates participated in this study, with 146 males (42.94%) and 194 females (57.06%). The participants were coming from two institutions: West China Hospital of Sichuan University (n = 134) and The First Affiliated Hospital of Chengdu Medical College (n = 206). In terms of academic standing, 206 students were in Year 1, 81 in Year 2, and 53 in Year 3. The average age of participants was 23.2 years (SD = 2.2). In China, postgraduate medical education follows a dual-track system, with clinical degrees (M.D./M.Med.) geared toward clinical practice and academic degrees (Ph.D./M.Sc.) aimed at research careers. Among the participants, 150 students (44.12%) were pursuing academic degrees, while 190 (55.88%) were enrolled in clinical degree programs across the two universities included in the study. Research areas included clinical medicine (n = 264, 77.65%), nursing science (n = 27, 7.94%), medical technology (n = 26, 7.65%), and oral medicine/dentistry (n = 23, 6.76%) (Table 1 ). Table 1 The demographic information of the studied medical students in this study, China 2025 Characteristics of participants The participants (n = 340) No. % University WCH of Sichuan University 134 39.41 1st Affiliated Hospital of Chengdu Medical College 206 60.59 Age (years) Mean ± SD 23.22 ± 2.20 Age Categories ≤ 23 years 181 53.24 >23 years 159 46.76 Sex Male 146 42.92 Female 194 57.06 Grade First year 206 60.59 Second year 81 23.82 Third year 53 15.59 Major Clinical medicine 264 77.65 Nursing science 27 7.94 Medical technology 26 7.65 Oral medicine/dentistry 23 6.76 Degree Academic degrees (Ph.D./M.Sc.) 150 44.12 Categories Clinical degrees (M.D./M.Med.) 190 55.88 Abbreviations: SD, Standard Deviation; Ph.D., Doctor of Philosophy; M.Sc., Master of Science; M.D., Doctor of Medicine; M.Med., Master of Medicine. Awareness, usage patterns and purposes of AI chatbot We found that 82.89% of the students demonstrated a solid understanding of and familiarity with general AI concepts. The majority (253 students, 74.6%) reported knowledge of machine learning or deep learning, while 210 students (61.95%) were familiar with natural language processing (NLP). Notably, only 76 students (22.35%) indicated familiarity with convolutional neural networks (CNNs). In terms of generative AI tool usage in daily academic activities, DeepSeek was the most widely used, with 90.86% of students reporting regular use. Over half of the respondents (55.16%) also used ChatGPT. Other AI tools were used to a lesser extent, including Gemini (9.14%), MetaAI (5.9%), Grok (4.13%), Co-Pilot (4.13%), and Claude (3.24%). When asked to identify their preferred AI chatbot, 246 (72.35%) of students selected DeepSeek as their top choice as illustrated in Fig. 1 . Regarding the frequency of AI chatbot usage, 33.53% of students reported using AI tools most of the time, while 45.00% used them often. An additional 18.53% used AI tools barely, and 2.94% never used AI chatbots. In terms of application, the most common uses among medical students included literature research and review (61.47%), academic question answering and research inspiration (55.88%), course study or exam preparation (55.00%), and English translation or editing (54.12%). A substantial proportion also employed AI tools for experimental design or data analysis (48.53%) and academic writing (43.24%). A smaller percentage (14.71%) utilized AI to assist with grant or project proposal preparation. (Fig. 2 a). Further analysis examined the specific sections of academic writing where students applied generative AI. The highest reported usage was for drafting introductions (47.94%), followed by methods sections (46.76%) and abstract writing (46.18%). Usage was also notable in discussion sections and grammar refinement (44.71% in each), and results sections (39.12%). Citation suggestions were less commonly supported by AI tools, with 31.76% of students reporting use in this area (Fig. 2 b). Among students who used AI chatbots, 99.11% reported overall satisfaction, with a mean satisfaction score of 4.4 on a 5-point Likert scale (1 = very dissatisfied, 5 = very satisfied). Regarding the perceived suitability of AI in different types of medical courses, 46.45% of students believed that AI is best suited to assist with medical literature reading. Additionally, 23.08% indicated that AI is applicable to basic medical sciences, such as anatomy and pathology. A smaller proportion (13.02%) considered AI suitable for research methodology. Notably fewer students felt that AI was well suited for doctor–patient communication simulation (10.06%) and clinical skills training (7.40%) (Fig. 3 ). Perceptions of AI chatbots To explore how medical students perceive the value and impact of AI chatbots in their academic education and experience, this study evaluated participants' views on the effectiveness of these tools in supporting both educational and research activities. Specifically, students were asked to evaluate how AI chatbots had influenced their performance in tasks such as information retrieval, academic writing, and knowledge synthesis. A majority of students (70.80%) reported that AI significantly improved the efficiency of information retrieval. Additionally, 64.31% indicated that AI contributed to more precise language use in writing. Over half of the respondents agreed that AI enhanced writing quality (56.05%) and facilitated a clearer structure and organization of ideas (58.11%). Notably, only three students (0.88%) believed that AI had not improve their writing efficiency. Students also reported perceived improvements in specific skills following AI tool usage. Notably, 78.10% believed their information retrieval efficiency had improved, 50.59% felt they had enhanced critical thinking abilities with AI assistance, 57.99% reported increased confidence in clinical decision-making, and 47.34% believed AI had helped improve long-term knowledge retention. Attitudes, concerns and expectations toward AI in medical education Overall, medical students exhibited a positive attitude toward the integration of AI in medical education. In this study, 84.03% of students agreed that it is essential to incorporate AI into the medical curriculum. More than half (51.19%) believed that AI-driven learning would eventually replace traditional classroom instruction. A substantial majority (85.51%) indicated that integrating AI with clinical medical education would improve overall learning efficiency. Furthermore, 82.54% of students felt that AI enhances personalized instruction and enables timely feedback. Regarding AI-based scoring systems, 68.04% believed they could improve the accuracy and fairness of academic assessments. Additionally, 73.01% agreed that AI-powered clinical simulation provides a safe environment for developing clinical skills. In the context of simulated diagnosis and case-based teaching, 77.81% of students reported that AI tools helped them identify key information and make preliminary judgments more quickly and accurately. In terms of future plans, 83.73% of students indicated they would apply AI tools in future teaching or research, and 77.22% planned to use AI to assess their own and their peers’ learning progress. A further 79.29% expressed willingness to recommend AI tools to classmates or mentors, and 78.99% expressed interest in participating in AI-related medical research projects (Fig. 4). Despite these benefits, students expressed several concerns regarding the limitations of AI in medical education. Over one-third (36.28%) were worried that AI may reduce opportunities for communication and collaboration among peers. Concerns were also raised about an overemphasis on data skills at the expense of soft skills (39.23%), diminished personalized attention from instructors (33.03%), and the over-quantification of learning outcomes (37.17%). Additionally, students reported worries about the accuracy of AI outputs (40.41%), potential for academic misconduct such as plagiarism (49.55%), data privacy and security issues (45.43%), and reduced critical thinking development (37.75%) (Fig. 5 ). Finally, students identified several key areas where they hoped AI chatbots would improve. These included enhanced accuracy and professionalism of responses (89.68%), better support for staying updated with research trends and literature (75.22%), improved contextual understanding in multi-turn conversations (72.27%), stronger data privacy and security protection (63.42%), and faster, more stable response performance (55.16%). Sensitive analysis and Pearson correlation analysis Sensitivity analyses were performed to compare awareness of AI chatbots, usage patterns, perceptions, attitudes, and concerns across gender and degree types. Usage patterns differed by gender, with female students more likely to report frequent AI use compared to males ( p = 0.025). Female students also expressed greater concerns regarding the risks of AI use in general ( p < 0.001) as well as risks specific to medical education ( p = 0.0002). No significant differences were observed between male and female students in terms of awareness scores, educational purposes of AI use, or overall attitudes toward AI. When stratified by degree program, clinical-track postgraduate students who have clinical programs demonstrated higher awareness of AI compared with their counterparts in academic-track postgraduate students who have academic programs ( p = 0.001). Additionally, clinical-track students reported higher overall satisfaction with AI chatbots ( p = 0.003). In contrast, no significant differences were found between the two groups regarding usage frequency, educational purposes of AI, or concern levels. Pearson correlation analysis was conducted to examine the relationships among students' awareness, frequency of AI usage, perception, attitudes, and concerns regarding AI in medical education. As shown in Fig. 6 , awareness of AI was positively associated with both usage frequency (r = 0.273, p < 0.001) and perception score of AI (r = 0.393, p < 0.001). Attitudes toward AI in education and in academic work were also significantly correlated with perception and usage. Notably, attitudes toward AI in education and academic work were strongly associated (r = 0.661, p < 0.001), suggesting consistent views across educational and academic contexts. While attitudes and perceptions of AI were generally positively correlated, concern-related variables exhibited weaker or even negative correlations with awareness and usage. Specifically, concerns about AI usage and concerns in medical education did not significantly correlate with awareness (r = 0.030, − 0.003 respectively) or usage frequency (r = − 0.080, − 0.104), suggesting that higher exposure to AI does not necessarily reduce levels of concern. However, the two concern domains were strongly positively correlated with each other (r = 0.782, p < 0.001), indicating that individuals who are more concerned about AI usage in general also tend to express greater concern about its impact in medical education. Additionally, attitudes toward AI in both education and academic work were moderately positively correlated with both concern variables (r = 0.310 to 0.249, all p < 0.001), implying that even those with favorable attitudes may still harbor significant concerns. Discussion While generative AI gained considerable purposes in medical field such as teaching, education, and research for years [ 13 ], ChatGPT or DeepSeek has garnered significant attention due to its open-access availability and generate human-like text or videos. Postgraduate medical education in China has two types of medical programs which differs from many other countries [ 14 ]. Clinical-track students are required to engage in research, often with publication demands, while academic-track students are simultaneously expected to balance intensive scientific work with some level of clinical exposure [ 15 ]. This dual emphasis creates unique pressures and learning needs, making it essential to investigate how Chinese postgraduate medical students adopt generative chatbots, assess their attitudes and concerns, and explore their behavioral intentions toward AI across both academic and clinical pathways. This cross-sectional study reported the generative chatbot adoption patterns in Chinese postgraduate medical students, assess their attitudes and concerns, and explore behavioral intentions toward AI in both academic and clinical-track programs students, providing evidence tailored to China’s medical education landscape. The findings indicated that most students were familiar with the concept of generative AI and its applications, yet remained cautious about its limitations, particularly regarding output accuracy, potential academic misconduct, and issues of data privacy and security. AI technologies are increasingly recognized for their ability to transform medical education by providing personalized instruction, adapting content to individual learning needs, and supporting remote education through interactive, intelligent platforms. These capabilities contribute to greater learning efficiency and improved educational quality [ 9 ]. Generative AI chatbots enhance accessibility to medical knowledge by offering scalable, personalized learning opportunities, including simulated patient interactions and real-time feedback[ 16 ]. They can also function as virtual tutors, creating engaging and adaptive learning environments tailored to students’ pace and preferences [ 17 ]. In our study, we found that generative AI tools—particularly large language models such as DeepSeek and ChatGPT—have rapidly gained traction among Chinese medical students. These chatbots were widely employed for literature review, academic writing, and self-directed learning. Similar findings have been reported in previous cross-sectional studies, where students expressed willingness to integrate ChatGPT into their academic workflow [ 18 ]. Reported benefits included improved information retrieval efficiency, enhanced confidence in clinical decision-making, and greater support for critical thinking. For postgraduate students, they apply deeper research improve the efficiency in clearer structure and ideas when writing. Our study found that 84.03% of students believed that AI is essential to incorporate into the medical curriculum in future teaching or training. Notably, approximately 77 ~ 84% of students in our survey expressed intentions to actively integrate AI tools into their future study and research, with many also indicating a willingness to recommend these tools to peers and faculty. Beyond academic writing, the programming capabilities of generative AI have also been recognized as valuable in fields such as bioinformatics education. Prior work demonstrated that AI chatbots can assist beginners in coding by generating accurate scripts, identifying errors, and providing real-time corrections and explanations, thereby improving both efficiency and motivation[ 19 ]. Beyond academic writing, our study found that over two-thirds of students reported that AI tools helped them identify key information and make preliminary judgments more quickly and accurately during simulated diagnosis and case-based teaching. Acting as virtual tutors, generative AI tools provided personalized and interactive learning experiences tailored to each student's unique needs [ 20 ]. This underscores how AI technologies can enrich clinical learning by simulating diagnostic scenarios and bolstering learners’ analytical speed and precision[ 21 ]. Such virtual tutor systems adapt to individual performance, delivering immediate feedback and guiding students through complex clinical reasoning tasks—thereby enhancing both efficiency and educational quality[ 22 ]. AI chatbots can also facilitate clinical research reported in our survey. 77.81% of students reported that AI tools helped them identify key information and make preliminary judgments more quickly and accurately in the context of simulated diagnosis and case-based teaching. ChatGPT is able to help doctors resolve problems encountered in clinical practice and form research ideas. Previous study has showed that ChatGPT has been used to assist in generating research questions, refining clinical hypotheses, and exploring under-researched areas by synthesizing relevant literature and suggesting plausible study frameworks [ 8 ]. In a pilot evaluation, ChatGPT was able to propose clinically relevant study topics that aligned with current evidence and research priorities, demonstrating its utility as a brainstorming partner in the early stages of research design[ 6 ]. In medical training, Liaw et al. suggested that AI-powered doctors gained a significant improvement in sepsis care knowledge from baseline comparted with the human-controlled group [ 23 ]. While tools like DeepSeek are still emerging in the literature, their functionally similar language-generation capabilities suggest similar promise. In future studies, blending AI-powered training with traditional human-controlled is believed to optimize clinical performance in clinical training and interprofessional communication. Interestingly, in the current study, we found that postgraduate students in clinical-track programs reported significantly higher awareness scores of AI and perceived greater benefits from AI compared with those in academic-track programs. This difference may reflect the fact that clinical-track students encounter AI tools more directly in their training, where applications such as clinical decision support, diagnostic imaging, and patient management systems are increasingly integrated. In contrast, academic-track students may engage more with traditional research methodologies and thus have lower exposure to AI-assisted tools, leading to comparatively lower awareness. These findings highlight the importance of tailoring AI education to the specific needs and contexts of both clinical and academic postgraduate training pathways[ 20 , 24 ]. Despite these advantages, concerns remain regarding content accuracy, ethical use, and overreliance on AI tools in our study. Students noted that AI-generated content sometimes included factual inaccuracies, nonsensical statements, insufficient sourcing, or fabricated references[ 25 , 26 ]. Broader concerns raised in the literature include risks of plagiarism, disputes over authorship, copyright violations, and potential misuse of AI-generated material in academic assessments[ 9 , 27 ]. These issues highlight the importance of critical evaluation of AI outputs by students, alongside clear institutional policies to ensure academic integrity. In the current study, there was a statistically significantly difference between males and females in concerns towards AI where females had much more concern than males about the AI, although females used AI more frequency than males. This difference between males and females could be attribute to that in some contexts, female medical students may feel more pressure to achieve academic success and therefore use AI tools intensively, but at the same time they may be more sensitive to ethical and professional standards, which heightens concern. Studies in health informatics and e-learning show women tend to weigh potential harms more than men [ 28 ]. Taken together, the findings suggest that while generative AI holds considerable promise for advancing medical education, its integration should be accompanied by well-defined ethical guidelines, active supervision, and structured educational frameworks to ensure responsible and sustainable use. The outcomes of our study hold the view that students’ views on AI were associated with their familiar with AI theory and perceived advantages such as improved learning efficiency. The correlation analysis of our study reveals that medical students’ attitudes toward AI are significantly shaped by their awareness and usage frequency of AI tools. The presence of concern appears to function independently of awareness or usage. The lack of significant correlations between concern variables and AI exposure suggests that simply interacting more with AI does not diminish skepticism or unease. Previous research has shown that higher AI literacy—defined as both conceptual understanding and practical familiarity—correlates with more favorable attitudes among medical students and healthcare professionals[ 8 ]. However, increased exposure alone may not fully mitigate ethical concerns, such as data privacy, algorithmic bias, and loss of human interaction, which often persist even among AI users [ 29 ]. This aligns with findings from other studies indicating that while generative AI tools can enhance performance in tasks like writing and diagnosis simulation, skepticism remains, particularly around over-reliance and the erosion of critical thinking [ 30 ]. The coexistence of enthusiasm and caution in medical students reflects a nuanced stance: they are willing to adopt AI when its value is evident but remain alert to its limitations and long-term implications. Addressing this duality may require not only technical training but also structured dialogue around the role of AI in professional identity, ethics, and patient safety. This study has several limitations that should be considered when interpreting the findings. First, the sample was limited to postgraduate medical students from two universities in Sichuan, which may not fully represent medical students across China. Regional variation, educational resources, and personal backgrounds (e.g., urban–rural disparities, family environments, and undergraduate training) could influence perceptions and adoption of AI tools, thereby introducing potential sampling bias. Second, the cross-sectional design and modest sample size restrict the ability to assess causal relationships or long-term trajectories in AI adoption. Longitudinal and multi-center studies are needed to evaluate how medical students’ attitudes and practices toward generative AI evolve over time. Third, the reliance on self-reported data introduces the possibility of response bias, including recall errors and social desirability effects, which may have affected the accuracy of reported usage patterns and attitudes. Beyond these methodological constraints, it is important to note that AI in medical education remains a rapidly evolving field. These results in our study highlight a dual reality: medical students are eager to integrate AI into their learning and research, yet they remain cautious about its limitations. Given that AI chatbots in medical education is still in its exploratory pharases, future longitudinal and multi-institutional studies are needed to clarify the long-term impact of AI on educational outcomes and to develop evidence-based strategies that ensure AI is used responsibly, effectively, and equitably in medical training. Declarations Ethics approval and consent to participate This study was approved by the Institutional Review Board (IRB) of West China Hospital in Sichuan University (Approval No. 2023-1066). Participation was voluntary, and all participants were informed about the purpose of the study, procedures, and confidentiality measures. Written informed consent was obtained from all participants prior to data collection. Consent for publication Not applicable. This manuscript does not contain any individual person’s data in any form (including images, videos, or case reports). Availability of data and materials The datasets generated during the current study are not publicly available due to privacy restrictions concerning participant information but are available from the corresponding author on reasonable request. Competing interests The authors declare that they have no competing interests. Funding This study was supported by the following funding sources: Funder One, Sichuan University Higher Education Teaching Reform Project (Phase XI), Grant/Award Number: SCU11144 (Recipient: Lijun Zhao); Funder Two, Chengdu Science and Technology Bureau, Technological Innovation and Research & Development Project, Grant/Award Number: 2024-YF05-00409-SN (Recipient: Lijun Zhao); Funder Three, Science & Technology Department of Sichuan Province, Science and Technology Training Program, Industry Science Popularization Capacity Enhancement Project, Grant/Award Number: 2025JDKP0031 (Recipient: Lijun Zhao); Funder Four, West China Hospital, Sichuan University, 1·3·5 project for disciplines of excellence–Clinical Research Fund, Grant/Award Number: 2024HXFH020 (Recipient: Lijun Zhao); Funder Five, Chengdu Medical College, Education Reform Research Project, Grant/Award Number: JG201905 (Recipient: Xuegui Ju). The funding bodies had no role in the design of the study, data collection, analysis, interpretation of data, or in writing the manuscript. Author contributions Conceptualization: Lijun Zhao. Data curation: Lijun Zhao, Qianqian Han and Peijuan Li. Formal analysis: Lijun Zhao. Writing – original draft: Lijun Zhao. Writing – review and editing: Hua Dai and Xuegui Ju. Supervision: Hua Dai and Xuegui Ju. All authors actively participated in the research process, made substantial contributions to manuscript revisions, and carefully reviewed and approved the final version. All authors read and approved the final manuscript. Clinical trial number Not applicable. Acknowledgments None. References Liu X, Faes L, Kale AU, Wagner SK, Fu DJ, Bruynseels A, et al. A comparison of deep learning performance against health-care professionals in detecting diseases from medical imaging: a systematic review and meta-analysis. Lancet Digit Health. 2019;1(6):e271–97. Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44–56. guidance W. Ethics and governance of artificial intelligence for health. 2021. Abdelhafiz AS, Ali A, Maaly AM, Ziady HH, Sultan EA, Mahgoub MA. Knowledge, Perceptions and Attitude of Researchers Towards Using ChatGPT in Research. J Med Syst. 2024;48(1):26. Tolsgaard MG, Pusic MV, Sebok-Syer SS, Gin B, Svendsen MB, Syer MD, Brydges R, Cuddy MM, Boscardin CK. The fundamentals of Artificial Intelligence in medical education research: AMEE Guide 156. Med Teach. 2023;45(6):565–73. Kung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, Elepaño C, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLOS Digit Health. 2023;2(2):e0000198. Abdelhafiz AS, Farghly MI, Sultan EA, Abouelmagd ME, Ashmawy Y, Elsebaie EH. Medical students and ChatGPT: analyzing attitudes, practices, and academic perceptions. BMC Med Educ. 2025;25(1):187. Sallam M. ChatGPT Utility in Healthcare Education, Research, and Practice: Systematic Review on the Promising Perspectives and Valid Concerns. Healthc (Basel) 2023; 11(6). Duan S, Liu C, Rong T, Zhao Y, Liu B. Integrating AI in medical education: a comprehensive study of medical students' attitudes, concerns, and behavioral intentions. BMC Med Educ. 2025;25(1):599. Busch F, Hoffmann L, Truhn D, Ortiz-Prado E, Makowski MR, Bressem KK, Adams LC. Global cross-sectional student survey on AI in medical, dental, and veterinary education and practice at 192 faculties. BMC Med Educ. 2024;24(1):1066. Eysenbach G. Improving the quality of Web surveys: the Checklist for Reporting Results of Internet E-Surveys (CHERRIES). J Med Internet Res. 2004;6(3):e34. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344–9. Gordon M, Daniel M, Ajiboye A, Uraiby H, Xu NY, Bartlett R, et al. A scoping review of artificial intelligence in medical education: BEME Guide 84. Med Teach. 2024;46(4):446–70. Liu X, Feng J, Liu C, Chu R, Lv M, Zhong N, Tang Y, Li L, Song K. Medical Education Systems in China: Development, Status, and Evaluation. Acad Med. 2023;98(1):43–9. Tie H, Luo Y, Chen D. Thinkings on the reform of medical education system in China. Med Educ Online. 2024;29(1):2302677. Cook DA, Overgaard J, Pankratz VS, Del Fiol G, Aakre CA. Virtual Patients Using Large Language Models: Scalable, Contextualized Simulation of Clinician-Patient Dialogue With Feedback. J Med Internet Res. 2025;27:e68486. Ghorashi N, Ismail A, Ghosh P, Sidawy A, Javan R. AI-Powered Chatbots in Medical Education: Potential Applications and Implications. Cureus. 2023;15(8):e43271. Hu N, Jiang XQ, Wang YD, Kang YM, Xia Z, Chen HH, Duan SN, Chen DX. Status and perceptions of ChatGPT utilization among medical students: a survey-based study. BMC Med Educ. 2025;25(1):831. Shue E, Liu L, Li B, Feng Z, Li X, Hu G. Empowering beginners in bioinformatics with ChatGPT. Quant Biol. 2023;11(2):105–8. Banerjee M, Chiew D, Patel KT, Johns I, Chappell D, Linton N, et al. The impact of artificial intelligence on clinical education: perceptions of postgraduate trainee doctors in London (UK) and recommendations for trainers. BMC Med Educ. 2021;21(1):429. Almansour M, Soliman M, Aldekhyyel R, Binkheder S, Temsah MH, Malki KH. An Academic Viewpoint (2025) on the Integration of Generative Artificial Intelligence in Medical Education: Transforming Learning and Practices. Cureus. 2025;17(3):e81145. Cheng CT, Chen CC, Fu CY, Chaou CH, Wu YT, Hsu CP, et al. Artificial intelligence-based education assists medical students' interpretation of hip fracture. Insights Imaging. 2020;11(1):119. Liaw SY, Tan JZ, Bin Rusli KD, Ratan R, Zhou W, Lim S, Lau TC, Seah B, Chua WL. Artificial Intelligence Versus Human-Controlled Doctor in Virtual Reality Simulation for Sepsis Team Training: Randomized Controlled Study. J Med Internet Res. 2023;25:e47748. Li Q, Qin Y. AI in medical education: medical student perception, curriculum recommendations and design suggestions. BMC Med Educ. 2023;23(1):852. Gibney E. AI models fed AI-generated data quickly spew nonsense. Nature. 2024;632(8023):18–9. Shumailov I, Shumaylov Z, Zhao Y, Papernot N, Anderson R, Gal Y. AI models collapse when trained on recursively generated data. Nature. 2024;631(8022):755–9. Boscardin CK, Gin B, Golde PB, Hauer KE. ChatGPT and Generative Artificial Intelligence for Medical Education: Potential Impact and Opportunity. Acad Med. 2024;99(1):22–7. Homko CJ, Zamora L, Santamore WP, Kashem A, McConnell T, Bove AA. Gender differences in cardiovascular risk factors and risk perception among individuals with diabetes. Diabetes Educ. 2010;36(3):483–8. Malerbi FK, Nakayama LF, Gayle Dychiao R, Zago Ribeiro L, Villanueva C, Celi LA, Regatieri CV. Digital Education for the Deployment of Artificial Intelligence in Health Care. J Med Internet Res. 2023;25:e43333. Preiksaitis C, Rose C. Opportunities, Challenges, and Future Directions of Generative Artificial Intelligence in Medical Education: Scoping Review. JMIR Med Educ. 2023;9:e48785. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 18 Mar, 2026 Read the published version in BMC Medical Education → Version 1 posted Editorial decision: Revision requested 28 Nov, 2025 Reviews received at journal 14 Nov, 2025 Reviewers agreed at journal 11 Nov, 2025 Reviewers agreed at journal 24 Oct, 2025 Reviews received at journal 21 Oct, 2025 Reviewers agreed at journal 13 Oct, 2025 Reviews received at journal 07 Oct, 2025 Reviewers agreed at journal 07 Oct, 2025 Reviewers invited by journal 02 Oct, 2025 Editor invited by journal 10 Sep, 2025 Editor assigned by journal 09 Sep, 2025 Submission checks completed at journal 09 Sep, 2025 First submitted to journal 26 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-7461983","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":529520007,"identity":"86abf5cd-8f1a-4067-a235-eaa8151e7740","order_by":0,"name":"Lijun Zhao","email":"","orcid":"","institution":"General Practice Medical Center, West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Lijun","middleName":"","lastName":"Zhao","suffix":""},{"id":529520008,"identity":"80c53774-3dec-4dd9-a99a-d6b7d03fd26f","order_by":1,"name":"Qianqian Han","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Qianqian","middleName":"","lastName":"Han","suffix":""},{"id":529520009,"identity":"e2958c71-c80e-4b58-b3a4-5b426d4ccd96","order_by":2,"name":"Peijuan Li","email":"","orcid":"","institution":"General Practice Medical Center, West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Peijuan","middleName":"","lastName":"Li","suffix":""},{"id":529520010,"identity":"f3d8867f-9e48-4432-a05a-1543ec3f2c35","order_by":3,"name":"Hua Dai","email":"","orcid":"","institution":"General Practice Medical Center, West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Hua","middleName":"","lastName":"Dai","suffix":""},{"id":529520011,"identity":"161bc370-4058-4113-a710-f74c97ee8c29","order_by":4,"name":"Xuegui Ju","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA00lEQVRIiWNgGAWjYBACxgYIzcMPIh8wMBgQr0WyAchOIEYLHBgcIFYL84wcM4kPvw7LGB/vMX+QUGFjzMB++OgGvA6bkZYmObPvMI/ZmTOGDQln0swYeNLSbuDXknxMmrcHqOVGjmFDYtthGwYJIBu/lsQ26b9ALcYziNcCtIXhx2EeAwmIFjPCWnqeJVv2NqTzSJw5VjgD6BdjNkJ+MWzPMbzx44+1PX9784YPHypsDPvZDx/Dr6UBZFUbkggbPuUgIA8m/xBSNgpGwSgYBSMaAABWz0td+1nOnwAAAABJRU5ErkJggg==","orcid":"","institution":"The First Affiliated Hospital of Chengdu Medical College","correspondingAuthor":true,"prefix":"","firstName":"Xuegui","middleName":"","lastName":"Ju","suffix":""}],"badges":[],"createdAt":"2025-08-26 10:38:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7461983/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7461983/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12909-026-08947-9","type":"published","date":"2026-03-18T15:57:30+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":93639214,"identity":"5dc00fc3-c21c-4b2a-a0b9-2792b3afb1f3","added_by":"auto","created_at":"2025-10-16 02:06:02","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":824131,"visible":true,"origin":"","legend":"","description":"","filename":"2025.9.2Manuscriptbmcntplaintext.docx","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/7ba0487c59007edb4ddd80c9.docx"},{"id":93638456,"identity":"566b53cd-105f-4511-a350-e5458a8c816f","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7342,"visible":true,"origin":"","legend":"","description":"","filename":"5a69bb1e27ce4140b65cf5470e939d37.json","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/71f0e193472b6243c63479fc.json"},{"id":93638459,"identity":"beb8064c-6280-4333-9af2-2a10904522e1","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":99275,"visible":true,"origin":"","legend":"","description":"","filename":"5a69bb1e27ce4140b65cf5470e939d371enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/5dd6a9452718696e9423f8f9.xml"},{"id":93639763,"identity":"36fda95b-29f1-41b1-abc7-3be310e87c18","added_by":"auto","created_at":"2025-10-16 02:14:02","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35032,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/bcec40d72cce7938ffdca3f8.png"},{"id":93639213,"identity":"5a1d00c1-cb2d-43de-a592-d54d2d2e809d","added_by":"auto","created_at":"2025-10-16 02:06:02","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":37009,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/0d1a13fe7d54989c7d3f8d17.png"},{"id":93638467,"identity":"a0ade35c-4b78-4c8d-ade4-a3905d2b6505","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24488,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/269265ddc37a48c52b93603d.png"},{"id":93638465,"identity":"27bfb5c3-be2e-48ec-9bb3-d80b4becf6b2","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":76314,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/453299aa411b712747a09faf.png"},{"id":93638469,"identity":"d06f8036-bf31-479f-9023-83ef1b20eb02","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64052,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/fdb763211d2cebdb5b9b19f9.png"},{"id":93638471,"identity":"e448d771-3eba-4767-83f9-cd6a577f8eb3","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":28639,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/bd7c7f7ff4846337aba63174.png"},{"id":93639217,"identity":"3df6ab8d-331d-4885-99b4-d2dd0d7bc513","added_by":"auto","created_at":"2025-10-16 02:06:02","extension":"xml","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":96012,"visible":true,"origin":"","legend":"","description":"","filename":"5a69bb1e27ce4140b65cf5470e939d371structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/e6fc4f5a2bbc4def167ab9e2.xml"},{"id":93638472,"identity":"25459f70-a8b7-4bf6-a34b-e54ab3bd50e1","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"html","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":106119,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/1738bb4c6003c9899a1d8731.html"},{"id":93638457,"identity":"6a16c093-d4fe-40db-9143-d719d8105081","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":69511,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistribution of preferred AI chatbots among postgraduate medical students at two colleges in China.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/c199fe15d34211f251349d27.jpeg"},{"id":93638458,"identity":"444835ac-9516-4b86-84b4-8557251156ad","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":172710,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eApplications of AI chatbots among medical students.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a) Distribution of AI chatbot use across different academic and research activities, including literature review, academic support, course study, translation, experimental design, and writing. (b) Distribution of AI chatbot use in specific sections of academic writing, such as introductions, methods, abstracts, discussions, grammar refinement, results, and citation suggestions.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/9b37ad1296d5ef6b540ea325.jpeg"},{"id":93638461,"identity":"29a3c2c4-0362-4853-afaa-ad015be7a8a4","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":92533,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePerceived suitability of AI in different types of medical courses among postgraduate students.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe pie chart illustrates students’ views on the most appropriate applications of AI, including medical literature reading, basic medical sciences, research methodology, doctor–patient communication simulation, and clinical skills training.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/1bbc3acbeaab8ab96258f3ce.jpeg"},{"id":93639215,"identity":"d89a94c4-1ba8-46fc-9b04-a55fb2eaaa6c","added_by":"auto","created_at":"2025-10-16 02:06:02","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":522195,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAttitudes of postgraduate medical students toward the integration of AI in medical education.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/0bd8f40c33a35d3aa91ff170.jpeg"},{"id":93638468,"identity":"6a267bd9-913c-4c8d-aac7-74fcdf3e4902","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":509024,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConcerns of postgraduate medical students regarding the use of AI in medical education.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/6f40c6cb49550f59b396a5e2.jpeg"},{"id":93638464,"identity":"c4e63214-c712-4c5a-9b30-cd5f889b397c","added_by":"auto","created_at":"2025-10-16 01:58:02","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":90053,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation analysis of students’ awareness, usage, perceptions, attitudes, and concerns regarding AI in medical education. *p \u0026lt; 0.05, **p \u0026lt; 0.01, ***p \u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/a7f42bca971bbf76a225ec04.jpeg"},{"id":105223251,"identity":"3da9077f-0f72-40d6-997e-73314baf49b5","added_by":"auto","created_at":"2026-03-23 16:00:26","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2306459,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7461983/v1/49ed77ea-4c8d-4856-9e72-f23497451704.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Generative AI Chatbots in Chinese Postgraduate Medical Education: Adoption Patterns, Attitudes, and Concerns","fulltext":[{"header":"Introduction","content":"\u003cp\u003eArtificial Intelligence (AI) refers to computational systems capable of performing tasks that typically require human intelligence, encompassing domains such as machine learning (ML), natural language processing (NLP), and deep learning. In medicine, AI has progressed from theoretical models to clinically integrated tools, transforming fields such as medical imaging analysis (e.g., tumor detection in radiology), predictive diagnostics (e.g., sepsis risk stratification), drug discovery, and personalized treatment planning[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. The World Health Organization has acknowledged AI\u0026rsquo;s potential to address global healthcare disparities, particularly in resource-limited settings[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eGenerative AI\u0026mdash;an emerging subset of AI capable of producing novel text, images, or structured data\u0026mdash;represents a significant frontier with transformative implications for both clinical practice and medical education. Medical chatbots, powered by NLP, have traditionally been used to simulate human conversation, assist in scheduling, and manage electronic health records. More recently, generative AI chatbots such as ChatGPT and DeepSeek have demonstrated the ability to synthesize complex, context-aware responses to open-ended academic and clinical queries, thereby extending their applications beyond routine administrative tasks[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. AI-based educational tools, including intelligent virtual patient simulations, adaptive learning platforms, and generative question\u0026ndash;answer systems, are increasingly being integrated into medical curricula, with evidence suggesting improvements in teaching efficiency and student engagement[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe adoption of generative AI chatbots in medical education offers several advantages, including enhanced access to information, real-time language refinement, and personalized feedback[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. However, legitimate concerns persist, particularly regarding the generation of inaccurate information (\u0026ldquo;hallucinations\u0026rdquo;), the potential for plagiarism, biases in training data, and copyright violations[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Such issues may influence the perceptions, expectations, and attitudes of both students and educators, potentially shaping the effectiveness of AI-enhanced teaching[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e\u003cp\u003ePrevious research has suggested that postgraduate medical students may exhibit higher awareness of AI tools compared to undergraduates[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Nevertheless, systematic and comparative analyses of AI awareness and adoption across different phases of medical education remain limited. Moreover, although international studies have explored AI adoption in medical training[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], China\u0026rsquo;s medical education system\u0026mdash;distinguished by clear differences between clinical and academic degree pathways\u0026mdash;remains underexamined. Clinical postgraduate programs (M.D./M.Med.) emphasize hands-on hospital training across specialties, enabling graduates to transition directly into clinical practice with residency certification. In contrast, academic postgraduate programs (Ph.D./M.Sc.) focus on laboratory research, scientific publishing, and theoretical study, preparing graduates for careers in research or academia, but requiring additional clinical training for hospital-based roles. The fundamental distinction lies in training emphasis\u0026mdash;patient care versus research\u0026mdash;and the readiness for specific career pathways after graduation. How these differing forms of clinical exposure shape students\u0026rsquo; adoption of generative AI chatbots is not yet well understood. To address this gap, the present cross-sectional study investigates adoption patterns of generative chatbots among Chinese medical students, evaluates their attitudes and concerns, and explores behavioral intentions toward AI in both academic and clinical learning contexts, thereby providing evidence tailored to China\u0026rsquo;s unique medical education landscape.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eStudy design\u003c/h2\u003e\u003cp\u003eThis cross-sectional survey was conducted among postgraduate medical students from two universities in Chengdu, China, with the aim of assessing their knowledge, attitudes, concerns, and behavioral intentions toward the use of generative AI chatbots in academic and educational contexts. Data were collected between June and August 2025. The questionnaire was distributed to students across different specialties and stages of postgraduate training. Eligible participants were current postgraduate medical students who had access to electronic devices and consented to participate in the study. Exclusion criteria were applied to individuals who were not enrolled in medical universities, lacked internet access, or declined participation. This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board at the West China Hospital of Sichuan University. All participants provided informed consent prior to their inclusion in the study. Electronic consent was obtained from every participant in accordance with institutional and ethical guidelines.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eQuestionnaire content\u003c/h3\u003e\n\u003cp\u003eThe online survey was developed in accordance with the \u003cem\u003eChecklist for Reporting Results of Internet E-Surveys (CHERRIES)\u003c/em\u003e[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] and followed the \u003cem\u003eStrengthening the Reporting of Observational Studies in Epidemiology (STROBE)\u003c/em\u003e guidelines[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. The final questionnaire comprised 40 items organized into seven sections and was administered via the WJX platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.wjx.cn/\u003c/span\u003e\u003cspan address=\"https://www.wjx.cn/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The survey addressed the following domains: participant demographics, awareness and understanding of AI, patterns of AI chatbot use, educational applications of AI, perceptions of generative AI, and students\u0026rsquo; attitudes, concerns, and expectations regarding AI in medical education and academic research. The survey link was disseminated through community engagement groups on the WeChat social media platform (Tencent Inc., Shenzhen, China).\u003c/p\u003e\n\u003ch3\u003eData cleaning and quality control\u003c/h3\u003e\n\u003cp\u003eTo ensure data integrity and reliability, responses underwent a structured cleaning process. Exclusion criteria were applied as follows: (1) responses with excessive missing data, defined as more than one unanswered item; (2) submissions demonstrating uniform or highly repetitive answer patterns suggestive of inattentive responding; and (3) completion times shorter than 50 seconds for the entire questionnaire or under 30 seconds after answering the first 11 questions. These thresholds were established through a pilot test involving five medical students from the research team, who determined benchmarks for reasonable completion times. A total of 345 students participated, of which 5 responses met the exclusion criteria and were removed from the final dataset.\u003c/p\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eSurvey data were exported from the WJX platform into Microsoft Excel for initial management and subsequently analyzed using Stata software (version 17.0, StataCorp, College Station, TX, USA). Descriptive statistics were used to summarize the data: categorical variables were presented as frequencies and percentages, and continuous variables as means with standard deviations (SD). Independent samples \u003cem\u003et\u003c/em\u003e-tests were applied to compare continuous variables between groups, while chi-square (χ\u0026sup2;) tests were used for categorical variables. Pearson correlation analysis was conducted to explore relationships among students\u0026rsquo; awareness, frequency of AI usage, perceptions, attitudes, and concerns regarding AI in medical education. A two-tailed p-value of \u0026lt;\u0026thinsp;0.05 was considered indicative of statistical significance.\u003c/p\u003e\u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eParticipant demographics\u003c/h2\u003e\u003cp\u003eA total of 340 medical postgraduates participated in this study, with 146 males (42.94%) and 194 females (57.06%). The participants were coming from two institutions: West China Hospital of Sichuan University (n\u0026thinsp;=\u0026thinsp;134) and The First Affiliated Hospital of Chengdu Medical College (n\u0026thinsp;=\u0026thinsp;206). In terms of academic standing, 206 students were in Year 1, 81 in Year 2, and 53 in Year 3. The average age of participants was 23.2 years (SD\u0026thinsp;=\u0026thinsp;2.2).\u003c/p\u003e\u003cp\u003eIn China, postgraduate medical education follows a dual-track system, with clinical degrees (M.D./M.Med.) geared toward clinical practice and academic degrees (Ph.D./M.Sc.) aimed at research careers. Among the participants, 150 students (44.12%) were pursuing academic degrees, while 190 (55.88%) were enrolled in clinical degree programs across the two universities included in the study. Research areas included clinical medicine (n\u0026thinsp;=\u0026thinsp;264, 77.65%), nursing science (n\u0026thinsp;=\u0026thinsp;27, 7.94%), medical technology (n\u0026thinsp;=\u0026thinsp;26, 7.65%), and oral medicine/dentistry (n\u0026thinsp;=\u0026thinsp;23, 6.76%) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eThe demographic information of the studied medical students in this study, China 2025\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003eCharacteristics of participants\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003eThe participants (n\u0026thinsp;=\u0026thinsp;340)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNo.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUniversity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWCH of Sichuan University\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e134\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e39.41\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1st Affiliated Hospital of Chengdu Medical College\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e206\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e60.59\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge (years)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e23.22\u0026thinsp;\u0026plusmn;\u0026thinsp;2.20\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge Categories\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026le;\u0026thinsp;23 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e181\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e53.24\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026gt;23 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e159\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e46.76\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSex\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e146\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e42.92\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e194\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e57.06\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGrade\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFirst year\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e206\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e60.59\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSecond year\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e81\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e23.82\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eThird year\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e53\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e15.59\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMajor\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eClinical medicine\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e264\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e77.65\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNursing science\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e7.94\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMedical technology\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e7.65\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOral medicine/dentistry\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e23\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e6.76\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDegree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAcademic degrees (Ph.D./M.Sc.)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e150\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e44.12\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCategories\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eClinical degrees (M.D./M.Med.)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e190\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e55.88\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"4\"\u003eAbbreviations: SD, Standard Deviation; Ph.D., Doctor of Philosophy; M.Sc., Master of Science; M.D., Doctor of Medicine; M.Med., Master of Medicine.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eAwareness, usage patterns and purposes of AI chatbot\u003c/h3\u003e\n\u003cp\u003eWe found that 82.89% of the students demonstrated a solid understanding of and familiarity with general AI concepts. The majority (253 students, 74.6%) reported knowledge of machine learning or deep learning, while 210 students (61.95%) were familiar with natural language processing (NLP). Notably, only 76 students (22.35%) indicated familiarity with convolutional neural networks (CNNs). In terms of generative AI tool usage in daily academic activities, DeepSeek was the most widely used, with 90.86% of students reporting regular use. Over half of the respondents (55.16%) also used ChatGPT. Other AI tools were used to a lesser extent, including Gemini (9.14%), MetaAI (5.9%), Grok (4.13%), Co-Pilot (4.13%), and Claude (3.24%). When asked to identify their preferred AI chatbot, 246 (72.35%) of students selected DeepSeek as their top choice as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003eRegarding the frequency of AI chatbot usage, 33.53% of students reported using AI tools most of the time, while 45.00% used them often. An additional 18.53% used AI tools barely, and 2.94% never used AI chatbots. In terms of application, the most common uses among medical students included literature research and review (61.47%), academic question answering and research inspiration (55.88%), course study or exam preparation (55.00%), and English translation or editing (54.12%). A substantial proportion also employed AI tools for experimental design or data analysis (48.53%) and academic writing (43.24%). A smaller percentage (14.71%) utilized AI to assist with grant or project proposal preparation. (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea).\u003c/p\u003e\u003cp\u003eFurther analysis examined the specific sections of academic writing where students applied generative AI. The highest reported usage was for drafting introductions (47.94%), followed by methods sections (46.76%) and abstract writing (46.18%). Usage was also notable in discussion sections and grammar refinement (44.71% in each), and results sections (39.12%). Citation suggestions were less commonly supported by AI tools, with 31.76% of students reporting use in this area (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). Among students who used AI chatbots, 99.11% reported overall satisfaction, with a mean satisfaction score of 4.4 on a 5-point Likert scale (1\u0026thinsp;=\u0026thinsp;very dissatisfied, 5\u0026thinsp;=\u0026thinsp;very satisfied).\u003c/p\u003e\u003cp\u003eRegarding the perceived suitability of AI in different types of medical courses, 46.45% of students believed that AI is best suited to assist with medical literature reading. Additionally, 23.08% indicated that AI is applicable to basic medical sciences, such as anatomy and pathology. A smaller proportion (13.02%) considered AI suitable for research methodology. Notably fewer students felt that AI was well suited for doctor\u0026ndash;patient communication simulation (10.06%) and clinical skills training (7.40%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003ePerceptions of AI chatbots\u003c/h3\u003e\n\u003cp\u003eTo explore how medical students perceive the value and impact of AI chatbots in their academic education and experience, this study evaluated participants' views on the effectiveness of these tools in supporting both educational and research activities. Specifically, students were asked to evaluate how AI chatbots had influenced their performance in tasks such as information retrieval, academic writing, and knowledge synthesis. A majority of students (70.80%) reported that AI significantly improved the efficiency of information retrieval. Additionally, 64.31% indicated that AI contributed to more precise language use in writing. Over half of the respondents agreed that AI enhanced writing quality (56.05%) and facilitated a clearer structure and organization of ideas (58.11%). Notably, only three students (0.88%) believed that AI had not improve their writing efficiency.\u003c/p\u003e\u003cp\u003eStudents also reported perceived improvements in specific skills following AI tool usage. Notably, 78.10% believed their information retrieval efficiency had improved, 50.59% felt they had enhanced critical thinking abilities with AI assistance, 57.99% reported increased confidence in clinical decision-making, and 47.34% believed AI had helped improve long-term knowledge retention.\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eAttitudes, concerns and expectations toward AI in medical education\u003c/h2\u003e\u003cp\u003eOverall, medical students exhibited a positive attitude toward the integration of AI in medical education. In this study, 84.03% of students agreed that it is essential to incorporate AI into the medical curriculum. More than half (51.19%) believed that AI-driven learning would eventually replace traditional classroom instruction. A substantial majority (85.51%) indicated that integrating AI with clinical medical education would improve overall learning efficiency. Furthermore, 82.54% of students felt that AI enhances personalized instruction and enables timely feedback. Regarding AI-based scoring systems, 68.04% believed they could improve the accuracy and fairness of academic assessments. Additionally, 73.01% agreed that AI-powered clinical simulation provides a safe environment for developing clinical skills. In the context of simulated diagnosis and case-based teaching, 77.81% of students reported that AI tools helped them identify key information and make preliminary judgments more quickly and accurately. In terms of future plans, 83.73% of students indicated they would apply AI tools in future teaching or research, and 77.22% planned to use AI to assess their own and their peers\u0026rsquo; learning progress. A further 79.29% expressed willingness to recommend AI tools to classmates or mentors, and 78.99% expressed interest in participating in AI-related medical research projects (Fig.\u0026nbsp;4).\u003c/p\u003e\u003cp\u003eDespite these benefits, students expressed several concerns regarding the limitations of AI in medical education. Over one-third (36.28%) were worried that AI may reduce opportunities for communication and collaboration among peers. Concerns were also raised about an overemphasis on data skills at the expense of soft skills (39.23%), diminished personalized attention from instructors (33.03%), and the over-quantification of learning outcomes (37.17%). Additionally, students reported worries about the accuracy of AI outputs (40.41%), potential for academic misconduct such as plagiarism (49.55%), data privacy and security issues (45.43%), and reduced critical thinking development (37.75%) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eFinally, students identified several key areas where they hoped AI chatbots would improve. These included enhanced accuracy and professionalism of responses (89.68%), better support for staying updated with research trends and literature (75.22%), improved contextual understanding in multi-turn conversations (72.27%), stronger data privacy and security protection (63.42%), and faster, more stable response performance (55.16%).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eSensitive analysis and Pearson correlation analysis\u003c/h2\u003e\u003cp\u003eSensitivity analyses were performed to compare awareness of AI chatbots, usage patterns, perceptions, attitudes, and concerns across gender and degree types. Usage patterns differed by gender, with female students more likely to report frequent AI use compared to males (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.025). Female students also expressed greater concerns regarding the risks of AI use in general (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) as well as risks specific to medical education (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0002). No significant differences were observed between male and female students in terms of awareness scores, educational purposes of AI use, or overall attitudes toward AI.\u003c/p\u003e\u003cp\u003eWhen stratified by degree program, clinical-track postgraduate students who have clinical programs demonstrated higher awareness of AI compared with their counterparts in academic-track postgraduate students who have academic programs (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.001). Additionally, clinical-track students reported higher overall satisfaction with AI chatbots (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.003). In contrast, no significant differences were found between the two groups regarding usage frequency, educational purposes of AI, or concern levels.\u003c/p\u003e\u003cp\u003ePearson correlation analysis was conducted to examine the relationships among students' awareness, frequency of AI usage, perception, attitudes, and concerns regarding AI in medical education. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e, awareness of AI was positively associated with both usage frequency (r\u0026thinsp;=\u0026thinsp;0.273, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and perception score of AI (r\u0026thinsp;=\u0026thinsp;0.393, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Attitudes toward AI in education and in academic work were also significantly correlated with perception and usage. Notably, attitudes toward AI in education and academic work were strongly associated (r\u0026thinsp;=\u0026thinsp;0.661, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), suggesting consistent views across educational and academic contexts.\u003c/p\u003e\u003cp\u003eWhile attitudes and perceptions of AI were generally positively correlated, concern-related variables exhibited weaker or even negative correlations with awareness and usage. Specifically, concerns about AI usage and concerns in medical education did not significantly correlate with awareness (r\u0026thinsp;=\u0026thinsp;0.030, \u0026minus;\u0026thinsp;0.003 respectively) or usage frequency (r = \u0026minus;\u0026thinsp;0.080, \u0026minus;\u0026thinsp;0.104), suggesting that higher exposure to AI does not necessarily reduce levels of concern. However, the two concern domains were strongly positively correlated with each other (r\u0026thinsp;=\u0026thinsp;0.782, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), indicating that individuals who are more concerned about AI usage in general also tend to express greater concern about its impact in medical education. Additionally, attitudes toward AI in both education and academic work were moderately positively correlated with both concern variables (r\u0026thinsp;=\u0026thinsp;0.310 to 0.249, all p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), implying that even those with favorable attitudes may still harbor significant concerns.\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eWhile generative AI gained considerable purposes in medical field such as teaching, education, and research for years [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e], ChatGPT or DeepSeek has garnered significant attention due to its open-access availability and generate human-like text or videos. Postgraduate medical education in China has two types of medical programs which differs from many other countries [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Clinical-track students are required to engage in research, often with publication demands, while academic-track students are simultaneously expected to balance intensive scientific work with some level of clinical exposure [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. This dual emphasis creates unique pressures and learning needs, making it essential to investigate how Chinese postgraduate medical students adopt generative chatbots, assess their attitudes and concerns, and explore their behavioral intentions toward AI across both academic and clinical pathways. This cross-sectional study reported the generative chatbot adoption patterns in Chinese postgraduate medical students, assess their attitudes and concerns, and explore behavioral intentions toward AI in both academic and clinical-track programs students, providing evidence tailored to China\u0026rsquo;s medical education landscape. The findings indicated that most students were familiar with the concept of generative AI and its applications, yet remained cautious about its limitations, particularly regarding output accuracy, potential academic misconduct, and issues of data privacy and security.\u003c/p\u003e\u003cp\u003eAI technologies are increasingly recognized for their ability to transform medical education by providing personalized instruction, adapting content to individual learning needs, and supporting remote education through interactive, intelligent platforms. These capabilities contribute to greater learning efficiency and improved educational quality [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Generative AI chatbots enhance accessibility to medical knowledge by offering scalable, personalized learning opportunities, including simulated patient interactions and real-time feedback[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. They can also function as virtual tutors, creating engaging and adaptive learning environments tailored to students\u0026rsquo; pace and preferences [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. In our study, we found that generative AI tools\u0026mdash;particularly large language models such as DeepSeek and ChatGPT\u0026mdash;have rapidly gained traction among Chinese medical students. These chatbots were widely employed for literature review, academic writing, and self-directed learning. Similar findings have been reported in previous cross-sectional studies, where students expressed willingness to integrate ChatGPT into their academic workflow [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Reported benefits included improved information retrieval efficiency, enhanced confidence in clinical decision-making, and greater support for critical thinking. For postgraduate students, they apply deeper research improve the efficiency in clearer structure and ideas when writing. Our study found that 84.03% of students believed that AI is essential to incorporate into the medical curriculum in future teaching or training. Notably, approximately 77\u0026thinsp;~\u0026thinsp;84% of students in our survey expressed intentions to actively integrate AI tools into their future study and research, with many also indicating a willingness to recommend these tools to peers and faculty. Beyond academic writing, the programming capabilities of generative AI have also been recognized as valuable in fields such as bioinformatics education. Prior work demonstrated that AI chatbots can assist beginners in coding by generating accurate scripts, identifying errors, and providing real-time corrections and explanations, thereby improving both efficiency and motivation[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eBeyond academic writing, our study found that over two-thirds of students reported that AI tools helped them identify key information and make preliminary judgments more quickly and accurately during simulated diagnosis and case-based teaching. Acting as virtual tutors, generative AI tools provided personalized and interactive learning experiences tailored to each student's unique needs [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. This underscores how AI technologies can enrich clinical learning by simulating diagnostic scenarios and bolstering learners\u0026rsquo; analytical speed and precision[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Such virtual tutor systems adapt to individual performance, delivering immediate feedback and guiding students through complex clinical reasoning tasks\u0026mdash;thereby enhancing both efficiency and educational quality[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAI chatbots can also facilitate clinical research reported in our survey. 77.81% of students reported that AI tools helped them identify key information and make preliminary judgments more quickly and accurately in the context of simulated diagnosis and case-based teaching. ChatGPT is able to help doctors resolve problems encountered in clinical practice and form research ideas. Previous study has showed that ChatGPT has been used to assist in generating research questions, refining clinical hypotheses, and exploring under-researched areas by synthesizing relevant literature and suggesting plausible study frameworks [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. In a pilot evaluation, ChatGPT was able to propose clinically relevant study topics that aligned with current evidence and research priorities, demonstrating its utility as a brainstorming partner in the early stages of research design[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. In medical training, Liaw \u003cem\u003eet al.\u003c/em\u003e suggested that AI-powered doctors gained a significant improvement in sepsis care knowledge from baseline comparted with the human-controlled group [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. While tools like DeepSeek are still emerging in the literature, their functionally similar language-generation capabilities suggest similar promise. In future studies, blending AI-powered training with traditional human-controlled is believed to optimize clinical performance in clinical training and interprofessional communication.\u003c/p\u003e\u003cp\u003eInterestingly, in the current study, we found that postgraduate students in clinical-track programs reported significantly higher awareness scores of AI and perceived greater benefits from AI compared with those in academic-track programs. This difference may reflect the fact that clinical-track students encounter AI tools more directly in their training, where applications such as clinical decision support, diagnostic imaging, and patient management systems are increasingly integrated. In contrast, academic-track students may engage more with traditional research methodologies and thus have lower exposure to AI-assisted tools, leading to comparatively lower awareness. These findings highlight the importance of tailoring AI education to the specific needs and contexts of both clinical and academic postgraduate training pathways[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eDespite these advantages, concerns remain regarding content accuracy, ethical use, and overreliance on AI tools in our study. Students noted that AI-generated content sometimes included factual inaccuracies, nonsensical statements, insufficient sourcing, or fabricated references[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Broader concerns raised in the literature include risks of plagiarism, disputes over authorship, copyright violations, and potential misuse of AI-generated material in academic assessments[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. These issues highlight the importance of critical evaluation of AI outputs by students, alongside clear institutional policies to ensure academic integrity. In the current study, there was a statistically significantly difference between males and females in concerns towards AI where females had much more concern than males about the AI, although females used AI more frequency than males. This difference between males and females could be attribute to that in some contexts, female medical students may feel more pressure to achieve academic success and therefore use AI tools intensively, but at the same time they may be more sensitive to ethical and professional standards, which heightens concern. Studies in health informatics and e-learning show women tend to weigh potential harms more than men [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Taken together, the findings suggest that while generative AI holds considerable promise for advancing medical education, its integration should be accompanied by well-defined ethical guidelines, active supervision, and structured educational frameworks to ensure responsible and sustainable use.\u003c/p\u003e\u003cp\u003eThe outcomes of our study hold the view that students\u0026rsquo; views on AI were associated with their familiar with AI theory and perceived advantages such as improved learning efficiency. The correlation analysis of our study reveals that medical students\u0026rsquo; attitudes toward AI are significantly shaped by their awareness and usage frequency of AI tools. The presence of concern appears to function independently of awareness or usage. The lack of significant correlations between concern variables and AI exposure suggests that simply interacting more with AI does not diminish skepticism or unease. Previous research has shown that higher AI literacy\u0026mdash;defined as both conceptual understanding and practical familiarity\u0026mdash;correlates with more favorable attitudes among medical students and healthcare professionals[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. However, increased exposure alone may not fully mitigate ethical concerns, such as data privacy, algorithmic bias, and loss of human interaction, which often persist even among AI users [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. This aligns with findings from other studies indicating that while generative AI tools can enhance performance in tasks like writing and diagnosis simulation, skepticism remains, particularly around over-reliance and the erosion of critical thinking [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. The coexistence of enthusiasm and caution in medical students reflects a nuanced stance: they are willing to adopt AI when its value is evident but remain alert to its limitations and long-term implications. Addressing this duality may require not only technical training but also structured dialogue around the role of AI in professional identity, ethics, and patient safety.\u003c/p\u003e\u003cp\u003eThis study has several limitations that should be considered when interpreting the findings. First, the sample was limited to postgraduate medical students from two universities in Sichuan, which may not fully represent medical students across China. Regional variation, educational resources, and personal backgrounds (e.g., urban\u0026ndash;rural disparities, family environments, and undergraduate training) could influence perceptions and adoption of AI tools, thereby introducing potential sampling bias. Second, the cross-sectional design and modest sample size restrict the ability to assess causal relationships or long-term trajectories in AI adoption. Longitudinal and multi-center studies are needed to evaluate how medical students\u0026rsquo; attitudes and practices toward generative AI evolve over time. Third, the reliance on self-reported data introduces the possibility of response bias, including recall errors and social desirability effects, which may have affected the accuracy of reported usage patterns and attitudes.\u003c/p\u003e\u003cp\u003eBeyond these methodological constraints, it is important to note that AI in medical education remains a rapidly evolving field. These results in our study highlight a dual reality: medical students are eager to integrate AI into their learning and research, yet they remain cautious about its limitations. Given that AI chatbots in medical education is still in its exploratory pharases, future longitudinal and multi-institutional studies are needed to clarify the long-term impact of AI on educational outcomes and to develop evidence-based strategies that ensure AI is used responsibly, effectively, and equitably in medical training.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthics approval and consent to participate\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Institutional Review Board (IRB) of West China Hospital in Sichuan University (Approval No. 2023-1066). Participation was voluntary, and all participants were informed about the purpose of the study, procedures, and confidentiality measures. Written informed consent was obtained from all participants prior to data collection.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eConsent for publication\u003c/p\u003e\n\u003cp\u003eNot applicable. This manuscript does not contain any individual person\u0026rsquo;s data in any form (including images, videos, or case reports).\u003c/p\u003e\n\u003cp\u003eAvailability of data and materials\u003c/p\u003e\n\u003cp\u003eThe datasets generated during the current study are not publicly available due to privacy restrictions concerning participant information but are available from the corresponding author on reasonable request.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eThis study was supported by the following funding sources: Funder One, Sichuan University Higher Education Teaching Reform Project (Phase XI), Grant/Award Number: SCU11144 (Recipient: Lijun Zhao); Funder Two, Chengdu Science and Technology Bureau, Technological Innovation and Research \u0026amp; Development Project, Grant/Award Number: 2024-YF05-00409-SN (Recipient: Lijun Zhao); Funder Three, Science \u0026amp; Technology Department of Sichuan Province, Science and Technology Training Program, Industry Science Popularization Capacity Enhancement Project, Grant/Award Number: 2025JDKP0031 (Recipient: Lijun Zhao); Funder Four, West China Hospital, Sichuan University, 1\u0026middot;3\u0026middot;5 project for disciplines of excellence\u0026ndash;Clinical Research Fund, Grant/Award Number: 2024HXFH020 (Recipient: Lijun Zhao); Funder Five, Chengdu Medical College, Education Reform Research Project, Grant/Award Number: JG201905 (Recipient: Xuegui Ju). The funding bodies had no role in the design of the study, data collection, analysis, interpretation of data, or in writing the manuscript.\u003c/p\u003e\n\u003cp\u003eAuthor contributions\u003c/p\u003e\n\u003cp\u003eConceptualization: Lijun Zhao. Data curation: Lijun Zhao, Qianqian Han and Peijuan Li. Formal analysis: Lijun Zhao. Writing \u0026ndash; original draft: Lijun Zhao. Writing \u0026ndash; review and editing: Hua Dai and Xuegui Ju. Supervision: Hua Dai and Xuegui Ju. All authors actively participated in the research process, made substantial contributions to manuscript revisions, and carefully reviewed and approved the final version. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003eAcknowledgments\u003c/p\u003e\n\u003cp\u003eNone.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLiu X, Faes L, Kale AU, Wagner SK, Fu DJ, Bruynseels A, et al. A comparison of deep learning performance against health-care professionals in detecting diseases from medical imaging: a systematic review and meta-analysis. Lancet Digit Health. 2019;1(6):e271\u0026ndash;97.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTopol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eguidance W. Ethics and governance of artificial intelligence for health. 2021.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbdelhafiz AS, Ali A, Maaly AM, Ziady HH, Sultan EA, Mahgoub MA. Knowledge, Perceptions and Attitude of Researchers Towards Using ChatGPT in Research. J Med Syst. 2024;48(1):26.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTolsgaard MG, Pusic MV, Sebok-Syer SS, Gin B, Svendsen MB, Syer MD, Brydges R, Cuddy MM, Boscardin CK. The fundamentals of Artificial Intelligence in medical education research: AMEE Guide 156. Med Teach. 2023;45(6):565\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, Elepa\u0026ntilde;o C, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLOS Digit Health. 2023;2(2):e0000198.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbdelhafiz AS, Farghly MI, Sultan EA, Abouelmagd ME, Ashmawy Y, Elsebaie EH. Medical students and ChatGPT: analyzing attitudes, practices, and academic perceptions. BMC Med Educ. 2025;25(1):187.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSallam M. ChatGPT Utility in Healthcare Education, Research, and Practice: Systematic Review on the Promising Perspectives and Valid Concerns. Healthc (Basel) 2023; 11(6).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDuan S, Liu C, Rong T, Zhao Y, Liu B. Integrating AI in medical education: a comprehensive study of medical students' attitudes, concerns, and behavioral intentions. BMC Med Educ. 2025;25(1):599.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBusch F, Hoffmann L, Truhn D, Ortiz-Prado E, Makowski MR, Bressem KK, Adams LC. Global cross-sectional student survey on AI in medical, dental, and veterinary education and practice at 192 faculties. BMC Med Educ. 2024;24(1):1066.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEysenbach G. Improving the quality of Web surveys: the Checklist for Reporting Results of Internet E-Surveys (CHERRIES). J Med Internet Res. 2004;6(3):e34.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evon Elm E, Altman DG, Egger M, Pocock SJ, G\u0026oslash;tzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGordon M, Daniel M, Ajiboye A, Uraiby H, Xu NY, Bartlett R, et al. A scoping review of artificial intelligence in medical education: BEME Guide 84. Med Teach. 2024;46(4):446\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu X, Feng J, Liu C, Chu R, Lv M, Zhong N, Tang Y, Li L, Song K. Medical Education Systems in China: Development, Status, and Evaluation. Acad Med. 2023;98(1):43\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTie H, Luo Y, Chen D. Thinkings on the reform of medical education system in China. Med Educ Online. 2024;29(1):2302677.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCook DA, Overgaard J, Pankratz VS, Del Fiol G, Aakre CA. Virtual Patients Using Large Language Models: Scalable, Contextualized Simulation of Clinician-Patient Dialogue With Feedback. J Med Internet Res. 2025;27:e68486.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGhorashi N, Ismail A, Ghosh P, Sidawy A, Javan R. AI-Powered Chatbots in Medical Education: Potential Applications and Implications. Cureus. 2023;15(8):e43271.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHu N, Jiang XQ, Wang YD, Kang YM, Xia Z, Chen HH, Duan SN, Chen DX. Status and perceptions of ChatGPT utilization among medical students: a survey-based study. BMC Med Educ. 2025;25(1):831.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShue E, Liu L, Li B, Feng Z, Li X, Hu G. Empowering beginners in bioinformatics with ChatGPT. Quant Biol. 2023;11(2):105\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBanerjee M, Chiew D, Patel KT, Johns I, Chappell D, Linton N, et al. The impact of artificial intelligence on clinical education: perceptions of postgraduate trainee doctors in London (UK) and recommendations for trainers. BMC Med Educ. 2021;21(1):429.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAlmansour M, Soliman M, Aldekhyyel R, Binkheder S, Temsah MH, Malki KH. An Academic Viewpoint (2025) on the Integration of Generative Artificial Intelligence in Medical Education: Transforming Learning and Practices. Cureus. 2025;17(3):e81145.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCheng CT, Chen CC, Fu CY, Chaou CH, Wu YT, Hsu CP, et al. Artificial intelligence-based education assists medical students' interpretation of hip fracture. Insights Imaging. 2020;11(1):119.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiaw SY, Tan JZ, Bin Rusli KD, Ratan R, Zhou W, Lim S, Lau TC, Seah B, Chua WL. Artificial Intelligence Versus Human-Controlled Doctor in Virtual Reality Simulation for Sepsis Team Training: Randomized Controlled Study. J Med Internet Res. 2023;25:e47748.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi Q, Qin Y. AI in medical education: medical student perception, curriculum recommendations and design suggestions. BMC Med Educ. 2023;23(1):852.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGibney E. AI models fed AI-generated data quickly spew nonsense. Nature. 2024;632(8023):18\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShumailov I, Shumaylov Z, Zhao Y, Papernot N, Anderson R, Gal Y. AI models collapse when trained on recursively generated data. Nature. 2024;631(8022):755\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBoscardin CK, Gin B, Golde PB, Hauer KE. ChatGPT and Generative Artificial Intelligence for Medical Education: Potential Impact and Opportunity. Acad Med. 2024;99(1):22\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHomko CJ, Zamora L, Santamore WP, Kashem A, McConnell T, Bove AA. Gender differences in cardiovascular risk factors and risk perception among individuals with diabetes. Diabetes Educ. 2010;36(3):483\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMalerbi FK, Nakayama LF, Gayle Dychiao R, Zago Ribeiro L, Villanueva C, Celi LA, Regatieri CV. Digital Education for the Deployment of Artificial Intelligence in Health Care. J Med Internet Res. 2023;25:e43333.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePreiksaitis C, Rose C. Opportunities, Challenges, and Future Directions of Generative Artificial Intelligence in Medical Education: Scoping Review. JMIR Med Educ. 2023;9:e48785.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-education","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"meed","sideBox":"Learn more about [BMC Medical Education](http://bmcmededuc.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/meed/default.aspx","title":"BMC Medical Education","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Generative artificial intelligence, Chatbots, Medical education, Technology adoption, Perceptions, Postgraduate medical students, China","lastPublishedDoi":"10.21203/rs.3.rs-7461983/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7461983/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eGenerative artificial intelligence (AI) chatbots are gaining attention in medical education for their potential to support academic writing, clinical reasoning, and personalized learning. However, little is known about their adoption, benefits, and concerns among Chinese postgraduate medical students, particularly across distinct clinical- and academic-track programs.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eA cross-sectional survey was conducted among 340 postgraduate medical students (146 males, 194 females) from two universities in Chengdu, China. A structured questionnaire assessed AI awareness, usage patterns, perceived benefits, attitudes, and concerns. Descriptive statistics, sensitivity analyses, and Pearson correlation analyses were applied to examine differences by gender and degree type.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eMost students (82.9%) reported strong AI awareness, with DeepSeek (90.9%) and ChatGPT (55.2%) being the most frequently used tools. Common applications included literature review (61.5%), exam preparation (55.0%), clinical case analysis (48.5%), and academic writing (43.2%). Reported benefits encompassed faster information retrieval (70.8%), improved writing precision (64.3%), and greater confidence in clinical decision-making (58.0%). Overall satisfaction was high (mean 4.4/5), and 84.0% of participants supported integration of AI into medical curricula. Female students reported more frequent use but greater concerns, while clinical-track students demonstrated higher awareness and satisfaction than academic-track students. Correlation analyses revealed positive associations among awareness, usage, perceptions, and attitudes, whereas concerns about accuracy, academic misconduct, and data security were largely independent of prior exposure.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eGenerative AI chatbots are widely adopted and valued by Chinese postgraduate medical students. Structured integration into medical curricula, accompanied by clear ethical safeguards, is essential to maximize benefits while addressing potential risks.\u003c/p\u003e","manuscriptTitle":"Generative AI Chatbots in Chinese Postgraduate Medical Education: Adoption Patterns, Attitudes, and Concerns","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-16 01:57:57","doi":"10.21203/rs.3.rs-7461983/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-11-28T07:42:07+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-15T03:55:55+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"137879584435089738158634726878314205720","date":"2025-11-11T08:14:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"202888092336195715320460906658724986932","date":"2025-10-24T11:42:24+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-21T14:22:53+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"188636472133179297332474925077464937321","date":"2025-10-13T12:07:21+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-07T06:56:36+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"325997395748362240052633266354590685143","date":"2025-10-07T06:00:56+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-10-02T10:42:04+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-09-10T17:08:55+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-09-09T06:30:00+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-09-09T06:29:54+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Education","date":"2025-08-26T10:25:36+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-education","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"meed","sideBox":"Learn more about [BMC Medical Education](http://bmcmededuc.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/meed/default.aspx","title":"BMC Medical Education","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c61ca044-1b32-444d-9447-796f611e32d0","owner":[],"postedDate":"October 16th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-03-23T16:00:01+00:00","versionOfRecord":{"articleIdentity":"rs-7461983","link":"https://doi.org/10.1186/s12909-026-08947-9","journal":{"identity":"bmc-medical-education","isVorOnly":false,"title":"BMC Medical Education"},"publishedOn":"2026-03-18 15:57:30","publishedOnDateReadable":"March 18th, 2026"},"versionCreatedAt":"2025-10-16 01:57:57","video":"","vorDoi":"10.1186/s12909-026-08947-9","vorDoiUrl":"https://doi.org/10.1186/s12909-026-08947-9","workflowStages":[]},"version":"v1","identity":"rs-7461983","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7461983","identity":"rs-7461983","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}