Artificial Intelligence for Medical Students: A Narrative Review and Practical Framework for Responsible AI Use in Undergraduate Medical Education (Preprint)

Abstract

<sec> <title>BACKGROUND</title> Artificial intelligence (AI), including large language models and other AI-enabled tools, are rapidly entering undergraduate medical education. However, practical guidance and formal teaching for safe and responsible use remains limited and inconsistent. </sec> <sec> <title>OBJECTIVE</title> This narrative review aims to synthesize current literature on AI use relevant to medical students and propose a practical framework for applications across preclerkship, clerkship, research, and future endeavors. </sec> <sec> <title>METHODS</title> We conducted a literature review of peer-reviewed and policy oriented literature on AI in medical education, emphasizing studies and guidance relevant to undergraduate medical education, generative AI, AI literacy, ethics, and student-facing workflows. Several themes were identified including foundational AI knowledge, educational use cases, clinical learning support, research applications, and ethical implementation. </sec> <sec> <title>RESULTS</title> Literature suggests that AI may support personalized learning, board preparation, clinical reasoning development, literature synthesis, and academic writing support. However, key risks include AI hallucination, automation bias, deskilling, privacy breaches, and biased outputs. A human-in-the-loop approach where AI is used as a supplement rather than a complete substitute is highly supported. We propose a practical framework for responsible student use centered on context-aware prompting, source verification, privacy protection, output validation, and transparent disclosure of AI technologies. </sec> <sec> <title>CONCLUSIONS</title> AI literacy is becoming an important competency in undergraduate medical education. Medical students should be trained not only to use AI tools effectively, but also to recognize their limitations, critically appraise outputs, and integrate them responsibly into their learning and future practice. </sec>

Ähnliche Arbeiten