Agentic education · Learning science

Actionable companion · Assessment & reasoning

Learning with agency: Status and perceptions of ChatGPT utilization among medical students: a survey-based study

AI-assisted editorial implementation guide, separate from the original research. Proposed actions are not claims or recommendations attributed to the source authors. No independent peer review of this guide is asserted.

Read original paperSearch paper agentCC BY 4.0

Evidence and context

Title and abstract screening; stage labels describe reported settings, not verified numerical ages. Design labels require full-text appraisal. Reported stage(s): Higher education. Health-professions learning must be distinguished from K–12 instruction.

Original authors’ abstract, reproduced under the linked source license:

Abstract Background The integration of ChatGPT with educational settings is happening at an unprecedented rate, and there is a growing trend for students to use ChatGPT for various academic work. Although numerous studies have evaluated the knowledge, attitudes, and practices related to ChatGPT among students in diverse medical fields, there remains a notable absence of such research within the Chinese context. Methods The questionnaire survey was conducted to a sample of 1,133 medical students from various medical colleges across Sichuan Province, China, between May 2024 and November 2024 to explore the awareness and attitudes of medical students towards ChatGPT. Descriptive statistics were used to tabulate the frequency of each variable. A chi-square test and multiple regression analysis were employed to investigate the factors influencing participants’ positive attitudes toward the prospective use of ChatGPT. Results The findings revealed that 62.9% of participants had employed ChatGPT in their medical studies, with 16.5% having utilized the tool in a published article. Participants primarily used ChatGPT for searching information (84.4%) and completing academic assignments (60.4%). However, concerns were expressed regarding the potential for ChatGPT to disseminate misinformation (76.9%) and facilitate plagiarism or complicate its detection (65.4%). Despite these concerns, 64.4% of respondents indicated a willingness to use ChatGPT to seek assistance with learning problems. Overall, a majority of participants (60.7%) maintained a positive attitude on the future use of ChatGPT in the medical field. Conclusion Our research showed that while most medical students perceived ChatGPT as a valuable tool for academic study and research, they remained cautious about its potential risks, particularly regarding misinformation and plagiarism concerns. Despite these reservations, a majority participants indicated a willingness to incorporate ChatGPT into their academic workflow, specifically for problem-solving tasks, and maintained optimistic perspectives regarding its potential integration into medical education and clinical practice. It is therefore essential to improve student literacy about AI, develop clear guidelines for its acceptable use, and implement support systems to ensure that medical students are fully prepared for the upcoming integration of AI into medical education. Trial registration Not applicable. Supplementary Information The online version contains supplementary material available at 10.1186/s12909-025-07438-7. Keywords: Artificial intelligence, ChatGPT, Medical education, Medical students

What kind of evidence is this?

Supporting AI learning science · Survey / observational study

Stage: Higher education. Screening labels describe reported context and study-design language. They are not a risk-of-bias rating or proof of effectiveness for an age band.

Study methods · source passage

Data from survey respondents remained anonymous, with no personal information collected. The study was deemed exempt research by the West China Second Hospital of Sichuan University Institutional Review Board (Approval No. 2024-083, Date: June 12, 2024) and used password-protected survey platform access for data security. The survey focused on Sichuan province-based medical students. Survey participation was voluntary and uncompensated.

Limitations and uncertainty · source passage

Several methodological limitations warrant consideration in interpreting our findings. Although convenience sampling facilitated efficient provincial data collection, this approach potentially introduced selection bias, with overrepresentation of participants from tertiary medical institutions possessing advanced educational resources. Consequently, medical colleges in prefecture-level cities, which frequently face substantial constraints in artificial intelligence infrastructure and clinical data accessibility, may be inadequately represented. Furthermore, the study encountered participation challenges among senior medical students engaged in clinical rotations, resulting in potentially reduced response rates from this survey. The AI literacy requirements and utilization patterns among students in non-clinical specialties, including public health and basic medical sciences, remain insufficiently documented, thus constraining the generalizability of our findings across diverse medical education contexts. Additionally, self-selection bias may have influenced our results, as students harboring favorable attitudes toward AI technology might have demonstrated greater propensity to participate in the survey, potentially yielding overly optimistic outcome estimates.

Design the system before the pilot

Use this study to inform a learning mechanism or implementation question. It is not direct evidence that autonomous agents improve learning. Keep the proposed agent behavior separate from what the source actually tested.

  • Use common rubrics, blind scoring where feasible and a fresh unaided reasoning task. Document permissible assistance and avoid treating AI-detector scores as proof of misconduct.

A 90-day district pilot

  1. Days 1–15 · Appraise. Have the curriculum lead and a teacher read the methods and limitations. Record the study design, sample, comparator, outcomes and evidence gaps. Define a local question that the study can help inform.
  2. Days 16–30 · Design. Choose one disciplinary task. Require a first attempt, a documented AI consultation, a critique of the output and an independent explanation. Align the comparison group and scoring rubric with the same objective. Name an accountable owner and set the baseline, success threshold and budget before implementation.
  3. Days 31–60 · Test. Start with a small voluntary cohort, preserve an accessible non-AI route and require teacher review. Use synthetic or de-identified examples; approve data handling and tool terms through district processes.
  4. Days 61–90 · Decide. Blind-score reasoning and accuracy on a new task without AI. Distinguish model examination scores, student perceptions and actual learner gains; only the last directly tests educational benefit. Compare with baseline, document uncertainty and stop or revise if learning worsens, access is unequal or critical errors persist.

The board decision brief

Report the learning problem, evidence quality, participating population, total cost including review time, measured benefit, errors, access differences and remaining uncertainties. Make a dated recommendation to stop, revise or expand. A positive perception survey alone does not establish learning improvement.

Attribution and reuse

Na Hu; Xiao Qin Jiang; Yi Da Wang; Yan Ming Kang; Zhen Xia; Hao Hui Chen; Sai Nan Duan; Dong Xu Chen. Status and perceptions of ChatGPT utilization among medical students: a survey-based study. BMC Medical Education (2025-06-04). https://doi.org/10.1186/s12909-025-07438-7

Companion prepared 2026-10-30. Original authorship and DOI are preserved. The abstract is an attributed source extract; the implementation plan is new AI-assisted editorial material, offered under the same CC BY 4.0 license. This guide has not undergone independent peer review.