Healthcare AI doesn’t change patient outcomes by itself — it changes outcomes when clinicians trust it and act on it. That “”last mile”” depends on medical education content that is accurate, evidence-based, and pedagogically sound. As LLMs generate increasingly fluent clinical text, the governance question shifts from can AI produce this content? to who is accountable when it does?

Drawing on real-world experience deploying AI in continuing medical education — an industry training over one million clinicians annually — this session introduces the ART framework (Accountability, Rigor, Taste) for governing human-AI collaboration in medical content creation. Attendees will learn how to structure AI-assisted workflows that maintain scientific rigor, define accountability chains when AI contributes to educational materials, and identify failure modes before they reach clinicians.