AMA releases governance framework for agentic AI in healthcare

The American Medical Association has released a governance framework for agentic AI used in healthcare, with four questions to guide how it assesses AI agents performing clinical tasks. The questions include what an agent is allowed to do and who is responsible for it. The AMA said the framework will guide its work with other organizations on using AI agents safely and responsibly in patient care.

The guidance covers tools used for ambient documentation, diagnostics and patient monitoring. Agentic AI can complete complex, multi-step tasks autonomously, which could allow inaccurate or biased decisions to cause patient harm before a clinician intervenes. In a survey of 250 U.S. healthcare leaders cited by the outlets, about 28% said their organization had deployed agentic AI tools and 44% said they were running pilots or proof-of-concept projects. 72% said tools were sometimes implemented without IT approval.