Bridging the gap between AI and traditional teaching methods in medical education: opportunities, challenges, and solutions

The integration of artificial intelligence (AI) into medical education has given rise to the innovative "AI + X" model, in which X refers to core disciplines such as clinical medicine, pharmacy, and nursing. Although this model enables personalized learning and immersive practical training, its widespread implementation also raises urgent ethical challenges. This study identifies key risks in AI-enabled medical education, including algorithmic bias that may shape students' professional perceptions and data privacy breaches resulting from unauthorized data use. A critical contribution of this study is the clear differentiation between generative AI tools (e.g., ChatGPT) and clinically validated AI systems (e.g., FDA-cleared radiology diagnostic tools), a distinction often overlooked in current curricula. The paper further examines the risks of using AI as a decision-support or moral guide, which may weaken humanistic values and professional autonomy, and discusses how superficial AI use may misalign with educators' roles, impair the patient-physician relationship, and foster student over-reliance. To address these challenges, we propose strengthening oversight of data processing and storage, improving transparency in AI decision-making, preserving human agency, fostering emotional engagement, enhancing educators' integrated teaching competencies, and cultivating students' critical thinking. Ultimately, the responsible integration of AI into medical education requires collaborative action from all stakeholders and should be guided by a three-dimensional ethical governance framework encompassing technological ethics, educational philosophy, and medical humanities.