Abstract: The transformative potential of artificial intelligence (AI) in medical education, characterized by its robust capabilities in data mining, image processing, and machine learning, is widely recognized. However, several challenges have been identified, including limited interpretability, questionable content accuracy, data security risks, cognitive biases among educators and students, and the absence of regulatory mechanisms. To address these concerns and ensure the sustainable development of AI-empowered medical education, a comprehensive study was conducted employing literature analysis and theoretical research.Based on stakeholder theory and collaborative governance theory, it systematically analyzes the core challenges of AI in medical education and explores the collaborative governance pathways among stakeholders such as engineers, medical experts/medical educators, students, and educational institutions. Effective governance was found to necessitate multi-stakeholder engagement across all implementation phases. During production, data governance must be strengthened to enhance AI interpretability and accuracy. For quality assurance, standardized evaluation systems should be established. Prior to deployment, strict reviews and rights protection measures are essential. In application, improving AI literacy among educators and students is critical. Ongoing regulation necessitates dynamic risk monitoring, while accountability mechanisms must clarify responsibility boundaries. Through the integration of technological innovation, institutional refinement, and stakeholder capacity-building, the convergence of medical education and modern technology can be achieved.

Key words: Artificial intelligence, Medical education, Stakeholders, Collaborative governance