In this response to a discussion paper and request for feedback issued by the Food and Drug Administration (FDA), scholars from Stanford HAI and Stanford University’s School of Medicine share reflections on the FDA’s considerations for the regulation of generative AI-enabled medical devices. Their recommendations focus on challenges related to operationalizing the two-axis risk assessment framework and ensuring that the proposed competency-based approach to evaluation and monitoring can both maximize patient safety and support innovation.

Stanford Health Policy's Michelle Mello, JD, PhD, a professor of health policy and of law, and Kevin Schulman, MD, a professor of medicine, health policy (by courtesy) and of operations, information and technology (by courtesy) at the Graduate School of Business, were among the co-authors of the paper that makes seven targeted recommendations. These address the FDA’s proposed two-axis risk framework—which weighs how autonomously a device acts against how much harm a wrong output could cause—as well as premarket and post-market evaluation.

Full Response and Paper