FDA Seeks Public Input on How to Regulate Generative AI in Medicine

The U.S. Food and Drug Administration is asking device manufacturers, clinicians, researchers, and the public how it should regulate generative AI in medicine. The agency's Digital Health Center of Excellence released a discussion paper on Tuesday addressing the unique risks and potential of generative AI-enabled medical devices, opening a formal comment period that runs through October 19, 2026. It is the latest regulatory move covered by AI Buzz Wire as governments worldwide grapple with fast-advancing medical AI.

The paper, first reported by trade outlets including Radiology Business and RAPS, is intended to engage stakeholders on the challenges and risks associated with the technology and to help advance future regulatory approaches at the FDA's Center for Devices and Radiological Health (CDRH).

What the FDA Is Asking

Generative AI-enabled medical devices hold "transformative promise" for patient care and the broader health ecosystem, the Digital Health Center of Excellence says. But the agency notes that these devices may introduce unique risks when compared to traditional software and other AI-enabled medical devices.

The discussion paper covers several topics and includes specific questions for each. The FDA is seeking feedback from device manufacturers, clinicians, researchers, and the public to help CDRH advance its understanding of the technology.

Importantly, the agency is lowering the barrier to participate: interested parties are not required to respond to every question and may provide partial responses based on their expertise, experience, or organizational capacity.

Discussion Paper, Not Guidance

The FDA has been careful to frame the document's legal status. The paper is for discussion purposes only and does not represent draft or final guidance, the agency emphasizes. It is not intended to propose or implement policy changes regarding how CDRH regulates these devices, nor does it communicate proposed regulatory expectations or requirements for supporting evidence in future marketing submissions.

The paper also does not address whether the approaches discussed fall within the FDA's existing legal authorities or whether new authorities would be necessary — a signal that the agency is still in an exploratory phase for this class of technology.

How to Comment

Comments are being accepted under docket number FDA-2026-N-7874 on Regulations.gov. The deadline for feedback is October 19, 2026.

The agency encourages participation from across the healthcare technology management sector to ensure a variety of perspectives are considered — an indication that the FDA wants input not only from large device makers but also from hospital biomedical engineering teams, health systems, and patient advocates.

Why Generative AI Is a Hard Case for Regulators

Generative AI poses a categorically different regulatory challenge from the locked-down, predetermined software the FDA historically reviewed. Traditional medical software produces predictable, testable outputs; generative systems can produce novel content, shift behavior with updates, and respond unpredictably to edge-case inputs.

For CDRH, the core questions include how to validate systems whose outputs cannot be exhaustively enumerated, how to handle post-market changes to model behavior, and what evidence developers should provide to demonstrate safety and effectiveness. The discussion paper's stakeholder questions aim to surface practical answers before the agency commits to a formal framework.

Healthcare AI adoption has accelerated sharply over the past two years, with generative tools moving from pilot projects to clinical documentation, imaging support, and decision-assist workflows. That pace has left regulators worldwide racing to keep up. The EU AI Act classifies medical AI as high-risk and imposes strict obligations on providers, while other jurisdictions are drafting frameworks of their own. Axios reported that the FDA is weighing approaches that would assess AI-enabled devices in ways analogous to how clinicians are evaluated — a reflection of how far the review paradigm may need to stretch.

What Happens Next

After the comment window closes on October 19, the FDA will typically review submissions and publish a summary of feedback. Depending on what it hears, the agency could move toward draft guidance, refine its premarket submission expectations, or propose new testing paradigms for generative features in devices.

For developers of medical AI, the docket is a rare early opportunity to shape the rules before they are written. Industry groups have historically pushed for flexibility on model updates and lifecycle management, while clinical safety advocates have urged rigorous post-market surveillance and transparency about model limitations.

The stakes are significant for patients too. Generative AI is increasingly embedded in tools that draft clinical notes, summarize patient histories, flag anomalies in scans, and suggest diagnoses. How the FDA draws the line between assistance and autonomy will help determine how quickly these tools reach clinics — and with what safeguards.

A Global Regulatory Moment

The docket opens amid a broader wave of regulatory activity around medical AI. Hospitals and health systems are deploying generative documentation and decision-support tools faster than standards bodies can write rules for them, and regulators from Europe to Asia are experimenting with frameworks that range from strict premarket review to post-market performance monitoring.

For the FDA, the challenge is compounded by the borderless nature of foundation models: the same underlying technology may power a consumer wellness chatbot, a hospital imaging assistant, and a drug-discovery pipeline, each carrying a very different risk profile. Discussion papers like this one are the agency's mechanism for mapping that terrain before committing to binding requirements.

Stakeholders who miss the October 19 deadline will have later opportunities — the FDA typically follows discussion papers with draft guidance and another comment window — but early input tends to carry the most weight in shaping the questions the agency prioritizes.

Stay Ahead of AI

Track the latest developments in AI regulation and policy at AI Buzz Wire, your source for daily AI industry coverage.

Read more AI news →