Written by Doug Wallace
Spoon Feed
A well-codified approach to AI research, regulatory oversight, and implementation practices is currently lacking. This JAMA summit report level-sets the key promises, pitfalls, and priorities for AI in medicine.
Smarter medicine (if we’re smart about it)
AI seems to be everywhere, and the health care landscape is no exception. AI scribes, radiology tools, clinical decision support systems, triage algorithms, and predictive analytics are gaining traction, with many more iterations on the horizon. This report summarizes perspectives from a 2023 JAMA-hosted summit of interdisciplinary experts who emphasized both the promise and pitfalls of AI adoption.
While AI offers potential to improve diagnostics, reduce clinician burden, and personalize care, concerns persist about bias, inaccuracies in output (i.e. hallucinations), motivations for use, and exacerbation of existing health inequities. The report stresses the need for clinical validation of AI tools, especially for high-risk applications, with well-supported patient and clinician involvement in AI development.
Panelists advocated for thoughtful AI deployment with regulatory oversight, continuous learning models, and robust research and implementation frameworks to ensure appropriate use.
How will this change my practice?
AI is here to stay, and it has the power to transform patient care for the better. What form it ultimately takes in on-the-ground health care delivery remains to be seen. We physicians have a key role to play in shaping the future of AI in medicine. I feel a responsibility, as a clinician and leader within my department, to ensure it is implemented safely and in a way that benefits providers and patients alike.
Source
AI, Health, and Health Care Today and Tomorrow: The JAMA Summit Report on Artificial Intelligence. JAMA. 2025 Oct 13. doi:10.1001/jama.2025.18490. Epub ahead of print. PMID: 41082366.
