The American Medical Association (AMA) rolled out a governance framework for agentic artificial intelligence solutions that undertake clinical functions.
The guide, created by the organization’s Center for Digital Health and AI, aims to ensure the technologies improve care while aligning clinical authority and accountability. The organization uses the term “augmented intelligence” when referring to AI to emphasize its assistive role in medicine.
The overarching theme of the framework is that clinical authority “should be earned through evidence and maintained through demonstrated performance.” It notes that authority and oversight can evolve alongside evidence, performance, risk and technological changes.
Based on this, the AMA says its work will focus on defining an AI system’s clinical role, validating its performance and safeguards, setting appropriate limits on its authority and establishing accountability for its use, including risks and outcomes.
The organization will use those focus areas to guide how it works with providers and stakeholders on liability, licensure and regulation approaches.
AMA CEO John Whyte, M.D., told Fierce Healthcare in an emailed statement that AI is “advancing toward taking on increasingly consequential clinical functions,” which is “raising urgent questions about clinical authority, oversight and accountability.”
“We cannot afford to wait for these questions to be answered through trial and error in patient care,” Whyte said. “The AMA's new framework sets out our ambitious agenda to address these questions, including the critical issues of liability, licensure and regulation. It will guide our work with physicians, patients, technology developers and policymakers to ensure clinical authority is earned through evidence, how oversight must evolve with risk, and how accountability is maintained.”
Whyte continued, “The time to do this work is now, while we still have the opportunity to shape how these technologies enter clinical practice.”
The agentic AI framework is the AMA’s latest effort to shape how AI is used and governed in healthcare.
In May, the organization released a comprehensive framework to protect physicians from unauthorized AI-driven deepfakes. It aims to close legal gaps while simultaneously modernizing physician identity protections.
The framework is guided by seven policy principles: physician identity as a protected right; prohibition on deceptive medical impersonation; informed, opt-in and revocable consent; mandatory transparency and labeling; shared responsibility to prevent impersonation; enforcement and practical remedies and minimizing administrative burden.
The organization also called on lawmakers in April to strengthen safeguards against AI chatbots amid increasing use for mental health services. Recommended safeguards include strict data protection standards, transparency standards and penalizing organizations for deceptive practices.