AMA Calls for Stronger FDA Oversight of Generative AI in Clinical Care

October 20, 2026 by matray

The American Medical Association has called on the U.S. Food & Drug Administration to establish a risk-based regulatory framework for generative artificial intelligence in medicine, emphasizing the need for clinical evidence, ongoing performance monitoring and manufacturer accountability as AI systems assume increasingly complex roles in patient care.

In an Oct. 19 letter to Rick Abramson, MD, director of the FDA’s Digital Health Center of Excellence, the AMA outlined recommendations for evaluating and regulating generative AI-enabled medical devices. The recommendations respond to an FDA discussion paper released Aug. 18 seeking public input on how the agency should oversee the emerging technology.

The AMA urged regulators to balance technological innovation with patient safety, warning that more flexible regulatory approaches must not weaken standards for demonstrating the safety and effectiveness of AI-enabled medical devices.

The organization identified two central principles for FDA oversight: Regulatory flexibility should not compromise patient safety, and physician involvement in evaluating AI technologies should not substitute for manufacturers demonstrating that their products perform safely and reliably.

The AMA recommended that regulatory requirements reflect a device’s intended clinical use, potential for patient harm and performance throughout its operational life. It also called for stronger post-market surveillance to identify problems that may emerge after AI systems enter clinical practice.

Unlike conventional medical software, generative AI systems can produce new content and perform increasingly sophisticated clinical functions, raising questions about how regulators should assess their reliability and monitor changes in performance over time.

The AMA emphasized that physicians should play a central role in developing clinical performance standards and identifying potential risks. However, it cautioned against relying on physician oversight as the primary safeguard against deficiencies in AI systems.

According to the organization, physician expertise should inform how AI performance is evaluated, but manufacturers must remain responsible for establishing that their products meet appropriate safety and effectiveness standards.

"Manufacturers must remain accountable for the performance of their products, and patients and physicians deserve the confidence that these technologies are safe and effective," said AMA president Willie Underwood, MD.

The AMA offered to collaborate with the FDA and other stakeholders to develop standards for physician evaluation of AI-generated clinical information, ongoing performance assessments and identification of AI failures that could affect patient care.

The recommendations come as the FDA considers new approaches to regulating generative AI-enabled medical devices. Its August discussion paper proposed a framework for assessing device risks, evaluating clinical performance before products reach the market and monitoring their safety after deployment.

The agency also sought feedback on the regulatory implications of increasingly autonomous AI systems capable of performing clinical tasks.

The AMA supports an approach that minimizes unnecessary regulatory burdens while maintaining safeguards proportionate to the risks posed by individual technologies.

The organization's recommendations also have implications for medical professional liability as AI systems take on more consequential clinical responsibilities, raising questions about the respective roles of physicians, technology developers and manufacturers when AI-generated information contributes to patient care.

The AMA has separately developed a governance framework addressing liability, licensure and regulation as increasingly capable AI systems assume clinical functions.

The organization said it intends to continue working with federal regulators, technology developers and physicians to establish standards that support innovation while protecting patients.