Toward a context-specific ethical framework for artificial intelligence in medical education: a critical narrative review informed by a systematic literature search

Abstract

Artificial intelligence (AI) is increasingly integrated into medical education, offering transformative opportunities for admissions, teaching, assessment, and curriculum design. However, the rapid adoption of AI introduces complex ethical challenges including transparency, fairness, accountability, data privacy, and the risk of over-reliance. Despite growing literature, there is limited synthesis of context-specific ethical frameworks tailored to medical education. This study aims to critically examine existing literature to identify, synthesize, and propose a context-specific ethical framework for AI integration in medical education. This study employed a critical narrative review design informed by a systematic literature search. A structured search of relevant databases was conducted to identify peer-reviewed articles published between 2019 and 2025 addressing ethical principles, challenges, and frameworks related to AI integration in medical and health professions education. Eligible studies included qualitative, quantitative, mixed-methods, cross-sectional, conceptual, perspective, viewpoint, review, and systematic review articles. Study selection was guided by a PRISMA-informed process, and the methodological quality of included studies was appraised using the SANRA and JBI critical appraisal tools. A critical narrative synthesis was subsequently undertaken to identify recurring ethical themes and develop a context-specific ethical framework. Analysis revealed five recurring domains of ethical concern: (1) Ethical foundations (transparency, fairness, accountability, social responsibility) (2), Educational integration and curriculum governance (3), Clinical training, simulation, and human oversight (4), Assessment, evaluation, and academic integrity, and (5) Data governance, privacy, and institutional responsibility. While many studies emphasize these principles, implementation strategies remain largely abstract, with minimal empirical validation and limited consideration of student perspectives, intersectional factors, or non-Western contexts. Recurrent challenges include algorithmic bias, inadequate faculty preparedness, data security, over-reliance on AI, and lack of standardized ethical oversight. Despite widespread recognition of AI’s potential in medical education, there is a pressing need for context-specific, empirically informed ethical frameworks. A five-domain layered framework derived from this review provides a structured approach to integrate ethical principles into curriculum design, clinical training, assessment, faculty development, and data governance. This framework aims to guide responsible AI use while preserving humanistic competencies, fostering transparency, fairness, accountability, and safeguarding learner and patient data.