FDA Weighs New Safety Rules for Generative AI Medical Devices - MyChesCo

The FDA has proposed a new regulatory framework for medical devices powered by generative artificial intelligence, focusing on how companies assess safety, prove clinical performance, and monitor products after patients begin using them. In a discussion paper, the agency outlined possible requirements covering risk classification, premarket evaluation, and postmarket surveillance. The FDA is taking public comments through October 19, 2026, and said generative AI can create risks different from conventional software and other AI-enabled devices. It also suggests a two-axis approach to risk assessment, and a competency-based premarket review combining nonclinical benchmarking with clinical confirmation.