JAMA +AI Weekly Update November 22, 2025

How can AI systems reason safely in the open world of medicine? JAMA+ AI Associate Editor Yulin Hswen, ScD, MPH, talks with Eric Horvitz, MD, PhD, Chief Scientific Officer at Microsoft, about the future of AI in 5 years to 100 years and how we can guide AI to be copilots while maintaining integrity and safety in the clinical arena.

They explore how decades of work in probabilistic modeling and human AI collaboration inform today’s research questions and what that means for authorship, trust, and innovation. The conversation looks backward and forward, exploring how AI can both diminish and rekindle independent thought.

Learn more and listen to the podcast.

Editor’s Picks in this week’s JAMA+ AI:   

• As generative AI tools expand beyond narrow use-cases, regulatory frameworks such as FDA’s SaMD may be insufficient. The authors of this Viewpoint propose adapting medical licensure principles—entry-to-practice standards, regular retesting, and disciplinary transparency—to govern clinical AI.
• Researchers at Vanderbilt University Medical Center evaluated an LLM–powered multiagent system for clinical decision support. Across 735 suggestions for 71 order sets, the system generated recommendations that were reviewed for accuracy, relevance, feasibility, and impact, with 19% judged as useful and 54% highly accurate.
• Could licensing medical AI as advanced practitioners provide flexible oversight compared to traditional FDA clearance? A clinician-style licensure system may better balance innovation and safety, tailoring supervision and liability based on the task and deployment context.