The fight over who regulates autonomous medical AI is happening at state level, and according to STAT it is being drafted in private. A 22 member workgroup convened by the Federation of State Medical Boards met in Washington in early October to develop guidelines for state medical boards, with executives from the American Medical Association, the American Hospital Association, Anthropic and insurer Elevance Health among the participants.
The focus, FSMB president Humayun Chaudhry told STAT, is AI tools that perform clinical functions with limited or no direct physician supervision. A draft could be made public in weeks to months, would be opened for feedback before adoption, and could form the basis for state legislation, since states individually regulate medical licensure and the practice of medicine.
The composition is the story. Doctor and hospital groups have historically argued that accountability must stay with a licensed clinician, a position the FSMB itself set out in its 2024 policy guidance. Technology companies and payers at the same table have commercial reasons to want a defined lane for autonomous systems, including reimbursement and liability clarity. Utah has already authorized an autonomous AI prescription pilot under a regulatory sandbox, which gives the workgroup a live experiment to react to.
Because the deliberations are confidential and the only account so far is STAT’s, the details here should be read as reported rather than confirmed. What is not in doubt is the direction: the first durable US rules for AI practicing medicine are more likely to arrive as state board guidance and state statutes than as a single federal framework, and the drafting has begun.