The American Medical Association has released a governance framework for agentic AI in healthcare, aiming to keep physicians central as autonomous tools spread into clinical work. The framework rests on the principle that AI agents performing clinical functions must be backed by evidence and maintain high-quality performance, and it poses four questions: what clinical function the AI performs, what demonstrates it can do that and what happens if it fails, what the AI should be permitted to do and under what conditions, and who is responsible for risks and outcomes.
The guidance arrives as agentic AI adoption accelerates. In a survey of 250 U.S. healthcare leaders, 28% said their organizations had already deployed such tools and 44% were running pilots. However, 72% said agentic AI is occasionally implemented without IT approval, and more than half cited compliance violations and unauthorized access to protected health information as major concerns.
The AMA's framework is intended to guide its work with other organizations, emphasizing that evidence, authority, oversight, and accountability must stay aligned. It comes amid a fragmented regulatory landscape: with no substantive federal AI regulation, states have introduced hundreds of AI-related healthcare bills, and industry groups are stepping in to provide guardrails. The source article does not compare the AMA framework to other proposals, so no independent assessment is offered here.{