FDA seeks feedback on framework for regulating GenAI devices
The US Food and Drug Administration (FDA) is seeking feedback on a competency-based approach to regulating medical devices that utilize generative artificial intelligence (GenAI).
This proposal was introduced in a discussion paper in which the agency seeks feedback on its proposed approach. The paper defines GenAI as “a class of AI models that emulate the structure and characteristics of input data in order to generate derived synthetic content. This can include images, videos, audio, text, and other digital content.”
Acting FDA Commissioner Kyle Diamantas said that “artificial intelligence is transforming medicine, and the United States must lead in shaping how this technology is developed and used safely and responsibly. Today’s announcement reflects the FDA’s commitment to advancing innovation for health care professionals and leveraging AI to improve care and patient health outcomes.”
“CDRH is considering a competency-based approach to premarket evaluation of GenAI enabled devices that is also inspired, at a high level, by how human clinicians are evaluated and credentialed, but would be adapted for the technical, practical, and legal considerations applicable to the regulation of medical devices,” FDA said.
The paper states that several studies have championed the use of models inspired by the evaluation and credentialing of human clinicians. FDA noted that two recent studies have concluded that “GenAI is too broad, adaptive, and opaque to be evaluated effectively through current medical device regulatory frameworks and instead propose competency-based regulation modeled on medical training, licensure examinations, supervised practice, periodic reevaluation, and public reporting.”
The competency approach would consist of non-clinical device benchmarking and clinical confirmation on how these devices perform. The paper states that “under such an approach, the final user-facing device, as configured and intended to be deployed for real-world use—and not the foundation model standing alone or other isolated subcomponent—would be evaluated.”
The paper also introduces a two-axis framework for assessing the risk posed by GenAI software functions; the horizontal axis represents the level of activity, scaling from informational (non-directive) to action-taking (fully autonomous), while the vertical axis represents the impact of the consequences, ranging from limited to severe.
According to an agency announcement, “GenAI-enabled medical devices hold transformative promise for patient care and the broader health ecosystem. At the same time, these devices may introduce unique risks when compared to traditional software and AI-enabled medical devices.”
The ultimate goal of a regulatory framework is to “enable a nimble regulatory approach that employs least burdensome principles and allows patients timely access to safe and effective medical devices.”
The agency has been considering a regulatory framework for AI-embedded medical devices since November 2024 when a public advisory committee meeting was held by the Center for Devices and Radiological Health (CDRH) with the Digital Health Advisory Committee (DHAC), where they addressed Total Product Life Cycle (TPLC) considerations for devices enabled by generative GenAI.
The paper also discusses some of the risks associated with the AI-embedded in medical devices. These include confabulations, or hallucinations, that may appear authentic to users, uncertainty in the bounds of a device’s intended use, limited visibility into underlying third-party foundation models, and performance degradation of the device and its components across test environments and in real-world applications across the total product life cycle.
The paper notes that the regulatory approach for software embedded in medical devices may not work for GenAI enabled medical devices.
“CDRH recognizes that evaluation approaches developed for software with bounded inputs and fixed outputs may not be appropriate for GenAI-enabled devices. Such evaluation has traditionally relied on extensive testing across a representative sample of device inputs and outputs; for GenAI-enabled devices, the range of possible inputs and outputs may be too large for such testing to be practical.”
The deadline for submitting comments is 19 October 2026. Submit comments to docket number FDA-2026-N-7874.
