FSMB: AI Not Ready for Independent Medical Licensing
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- The Federation of State Medical Boards contends that AI is not ready to be independently licensed like a physician, because medical licensure is a legal and ethical grant of authority tied to human judgment, professional accountability, and a duty to put patients first.
- FSMB's 2024 House of Delegates policy adopted nonbinding guidance encouraging physicians to use AI when it improves care and advising documentation of its use, while maintaining that professional responsibility — including liability for AI-related harms — remains with the licensed clinician.
- Idaho and Iowa each introduced legislation this year to create separate state licensing boards for "autonomous service providers"; both bills failed, though FSMB views the proposals as proof the question of licensing AI as clinicians is no longer hypothetical.
- Utah's Office of Artificial Intelligence Policy approved a 12-month pilot with health tech startup Doctronic to automate guideline-based prescription renewals for 192 commonly prescribed drugs under physician oversight — an arrangement that drew scrutiny from the state medical licensing board.
- FSMB has convened a Workgroup on the Regulation of AI in the Practice of Medicine and is updating its guidance to address the shift from passive-assistance AI toward agentic systems that act with limited or no direct physician supervision.
- The federation serves all 69 state and territorial medical boards and advises jurisdictions to reexamine how "the practice of medicine" is defined as AI systems grow more autonomous, positioning medical boards — not new AI-specific agencies — as the appropriate authorities for clinical AI oversight.
Why it matters: Because FSMB speaks for all 69 state and territorial medical boards, its stance sets the floor for how state regulators are likely to handle AI in clinical care: physicians — not machines or vendors — stay on the hook for patient harm. That keeps accountability clear but also means AI vendors and telehealth startups face a regulatory framework built around human licensees, complicating any business model that depends on autonomous AI acting independently of a clinician.
