The FDA has now cleared over 1,400 AI medical devices, and radiology owns three quarters of them
The regulator authorised more AI-enabled devices in 2025 than in any year on record. The numbers show where clinical AI is actually landing, and where it is not.
World Nexus GroupThe US Food and Drug Administration keeps a public register of every AI-enabled medical device it has authorised for marketing. By the end of December 2025 that list had passed 1,451 devices, cumulative since the agency began tracking them in the 1990s. 1
2025 was the busiest year the register has ever seen: 331 AI-enabled devices were authorised in a single year, more than any year before it. 1
Radiology is the centre of gravity
The headline figure hides how concentrated clinical AI still is. Roughly three quarters of all authorised AI devices are in radiology. That share has been remarkably stable, sitting around 75 to 76 percent across 2024 and 2025. 12
The reason is structural. Imaging produces large, standardised digital datasets with clear ground-truth labels, which is exactly the substrate machine learning needs. Specialties that run on messier, less structured data, primary care being the obvious example, have far fewer cleared devices despite far larger patient populations.
What this means for operators
Two things follow for anyone deploying AI inside a health system.
- The tooling for imaging-heavy pathways is mature and regulator-blessed. If your workflow touches radiology, cardiology imaging, or pathology, there is a real and growing shelf of cleared products to buy rather than build.
- Everywhere else, you are earlier than the headlines suggest. A "1,400 devices" number reads like saturation; in your specialty it may mean a handful.
The FDA also issued comprehensive draft guidance in 2025 for developers of AI-enabled devices, including expectations around lifecycle management and post-market monitoring, a signal that the regulator is shifting from one-off clearances toward how these tools behave once they are live in the clinic. 3
The practical takeaway: treat the device list as a map of where the evidence and the regulatory path already exist, not as a measure of how far AI has spread across medicine as a whole.
References
- 1.Artificial Intelligence-Enabled Medical Devices — official device list — U.S. Food and Drug Administration
- 2.FDA Updates AI List with New Clearances — The Imaging Wire
- 3.FDA Issues Comprehensive Draft Guidance for Developers of AI-Enabled Medical Devices — U.S. Food and Drug Administration
This briefing summarises publicly available research and reporting for information only. It is not medical, investment, or legal advice. Follow the references above to the primary sources.