Regulation and Ethical Frameworks for AI in Health Professions Education: A Critical Narrative Review
- 1Department of Medical Oncology, St. Vincent’s University Hospital, Dublin, Ireland
- 2School of Medicine, University College Dublin, Dublin, Ireland
- 3Independent Researcher, Dublin, Ireland
Abstract
Introduction:: Integrating artificial intelligence (AI)—particularly generative AI and large language models (LLMs)—into health professions education offers substantial pedagogical opportunities, while highlighting critical gaps in existing ethical and regulatory frameworks. This critical narrative review examines how contemporary literature conceptualizes and addresses these challenges.
Methods:: Publications from 2020–2025 were identified through searching PubMed, Scopus, and ERIC, focusing on AI, medical and health professions education, ethics, governance, and regulation. The initial search yielded 452 records; following two-stage screening (title/abstract and full-text review), thirty-four sources met the pre-defined inclusion criteria. These sources, including empirical studies, systematic reviews, policy documents, and conceptual papers, were included. Reflexive thematic analysis was used to synthesize the findings, while SANRA (Scale for the Assessment of Narrative Review Articles) informed the weighting of sources, which were categorized into high-, moderate-, and low-weight tiers.
Results:: Four themes emerged: 1) a regulatory vacuum marked by policy lag and fragmented institutional responses; 2) reinterpretation of autonomy, beneficence, nonmaleficence, and justice in relation to opaque, datadriven systems, 3) generative AI’s impact on scholarship, authorship, and academic integrity; and 4) equity, bias, and professional identity formation in AImediated learning environments. Integrating these themes with crosssector frameworks and the NIST AI Risk Management Framework, the review develops a riskbased governance taxonomy for AI in Health Professions Education, distinguishing low, medium and highrisk educational use cases.
Conclusion:: This synthesis indicates that current literature favors evolving beyond reactive policies. Consequently, we propose a theoretical shift toward risk-proportionate governance. We suggest that human-in-the-loop pedagogy should be prioritized as a conceptual framework to guide future research and policy development, rather than presenting it as a validated outcome.
Related
- Who Learns from Disruption? Antifragility and Burden Transfer in Health Professions Education
- Beyond the clicker: rethinking student response systems as pedagogical infrastructure for healthcare professions education
- Adapting the Girandole framework for cross-disciplinary social justice reflection in health professions education
- How to … Do Feasibility, Acceptability and Potential Effectiveness Studies in Health Professions Education Research
- Leveraging AI to Accelerate Competency-Based Education in the Health Professions.
BibTeX
@article{ahmed2026regulation,
title = {{Regulation and Ethical Frameworks for AI in Health Professions Education: A Critical Narrative Review}},
author = {YASAR AHMED and SIMAA KHAYAL},
journal = {Journal of Advances in Medical Education \& Professionalism},
year = {2026},
month = oct,
doi = {10.30476/jamp.2026.110257.2340},
url = {https://doi.org/10.30476/jamp.2026.110257.2340},
note = {CC BY-ND 4.0},
}
RIS
TY - JOUR
TI - Regulation and Ethical Frameworks for AI in Health Professions Education: A Critical Narrative Review
AU - YASAR AHMED
AU - SIMAA KHAYAL
PY - 2026
DA - 2026/10/01
JO - Journal of Advances in Medical Education & Professionalism
DO - 10.30476/jamp.2026.110257.2340
UR - https://doi.org/10.30476/jamp.2026.110257.2340
AB - Introduction:: Integrating artificial intelligence (AI)—particularly generative AI and large language models (LLMs)—into health professions education offers substantial pedagogical opportunities, while highlighting critical gaps in existing ethical and regulatory frameworks. This critical narrative review examines how contemporary literature conceptualizes and addresses these challenges. Methods:: Publications from 2020–2025 were identified through searching PubMed, Scopus, and ERIC, focusing on AI, medical and health professions education, ethics, governance, and regulation. The initial search yielded 452 records; following two-stage screening (title/abstract and full-text review), thirty-four sources met the pre-defined inclusion criteria. These sources, including empirical studies, systematic reviews, policy documents, and conceptual papers, were included. Reflexive thematic analysis was used to synthesize the findings, while SANRA (Scale for the Assessment of Narrative Review Articles) informed the weighting of sources, which were categorized into high-, moderate-, and low-weight tiers. Results:: Four themes emerged: 1) a regulatory vacuum marked by policy lag and fragmented institutional responses; 2) reinterpretation of autonomy, beneficence, nonmaleficence, and justice in relation to opaque, datadriven systems, 3) generative AI’s impact on scholarship, authorship, and academic integrity; and 4) equity, bias, and professional identity formation in AImediated learning environments. Integrating these themes with crosssector frameworks and the NIST AI Risk Management Framework, the review develops a riskbased governance taxonomy for AI in Health Professions Education, distinguishing low, medium and highrisk educational use cases. Conclusion:: This synthesis indicates that current literature favors evolving beyond reactive policies. Consequently, we propose a theoretical shift toward risk-proportionate governance. We suggest that human-in-the-loop pedagogy should be prioritized as a conceptual framework to guide future research and policy development, rather than presenting it as a validated outcome.
SN - 2322-2220
SN - 2322-3561
ER -