Request for Comments: AI-Enabled Medical Devices

On September 30, 2025, FDA released a Request for Public Comment: Measuring and Evaluating Artificial Intelligence-enabled Medical Device Performance in the Real-World. Comments are due to the docket by December 1, 2025.

The request for comments is to obtain feedback about the “current, practical approaches to measuring and evaluating the performance of AI-enabled medical devices in the real-world, including strategies for identifying and managing performance drift, such as detecting changes in input and output.” The driving concern appears to be assessing whether the devices “remain safe and effective throughout their life cycle.”

Specifically, the agency is looking for information on methods that are:

• Currently deployed at scale in real-world clinical environments,
• Supported by real-world evidence, and
• Applied in clinical (patient- or health care worker-facing) settings.

While the request sets forth a number of specific questions for consideration—many of which would appear to require disclosure of potentially confidential or proprietary information—a critical and unanswered question relates to how FDA intends to use the information it receives from this request. One paragraph in particular caught this blogger’s attention:

Currently, many AI-enabled medical devices are evaluated primarily through retrospective testing or static benchmarks. While these methods may help establish a baseline understanding of the medical device performance, they are not designed to predict behavior in dynamic, real-world environments. Ongoing, systematic performance monitoring is increasingly recognized as relevant to maintaining safe and effective AI use by observing how systems actually behave during clinical deployment.

While it may be too early to conclude that this language is cause for concern, it certainly should give manufacturers of AI-enabled devices a reason to take a pause and consider submitting comments.  Even assuming it is true that the data and information provided in a premarket submission are not designed or intended to demonstrate long-term safety and effectiveness, it is not clear how FDA intends to remedy this post-hoc. Would FDA try to impose post-market requirements on currently marketed AI-enabled devices even if those devices were authorized without any such requirements? What would be the legal authority for doing so? There could also be impacts on the pre-market side, for example, if FDA takes learnings from this public comment period to impose additional pre-market requirements moving forward that could potentially impact substantial equivalence determinations.

We agree that is it important to determine how to ensure the ongoing safety and effectiveness of AI-enabled devices, but in doing so it will be important not to impose new or different requirements for this particular device type, as that would require statutory changes. We would encourage manufacturers to submit comments emphasizing that post-market obligations for medical devices are appropriate for all device types, including AI-enabled device types, even if the methods for collecting that information may be different. We would be happy to help you with your submission.