Utilization of AI Among Medical Students and Development of AI Education Platforms in Medical Institutions: Cross-Sectional Study

Comprehensive Summary

This study, presented by Shi et al., examined how medical students utilize AI and their expectations for language models in an approach to compile systematic data to assess the why and how of artificial intelligence usage among med students. A cross sectional survey was administered anonymously to medical students studying in Shanghai, China between March 1 - May 31, 2025 which required questions on frequency of use, reason, and platform identification. Results indicated that the questionnaire was 97% effective with 428 valid participant answers of seven distinct medical schools. AI usage averaged 5 times a week on platforms namely, DeepSeek, ChatGPT, and Doubao. Undergraduate students were found to utilize such programs for exam help while graduate students focused on research assistance. Shi et al. acknowledged the lack of geographical diversity and risks of bias owning to self-reported data within the study.

Outcomes and Implications

Students, whether graduate or undergraduate, are already using AI to learn despite the one-size-fits-all approach in most available applications. However, users have highly specified requests and AI platforms have yet to develop features to accommodate these demands. Educational stage is a significant determinant of it’s usage and these tailored improvements could greatly increase its effectiveness. Not only that, but AI integration has been proposed for the healthcare system, medical students familiar with artificial intelligence applications will be that much more adjusted to these changes.