The FDA opened a consultation on how it should regulate generative AI-enabled medical devices, aiming to refine risk assessment, premarket review expectations, and postmarketing monitoring. The agency is targeting practical answers to how such systems should be evaluated when their outputs can change after deployment or evolve with data. For device makers and clinical innovators, the document shifts the compliance conversation from whether models work to how they should be governed across the product lifecycle—especially around safety, intended use, and measurement of performance in real-world settings. Industry watchpoints now include what documentation will be expected for premarket submissions and what monitoring and update mechanisms will qualify as sufficient post-authorization oversight.