FDA AI Medical Device Approval Statistics: 2026 Report

FDA AI Medical Device Approvals by Year

The yearly count shows how quickly AI has moved into cleared medical products. The Healthcare Industry Reviews research team counted each device by the year of its final FDA decision.

• The FDA authorized 335 AI-enabled devices in 2025, up 42.6% from 235 in 2024. The gain of 100 devices is the largest one-year increase on the list.
• From January through June 29, 2026, the FDA authorized 181 devices, against 173 from January through June 29, 2025, an increase of 4.6%.
• Nearly all new AI devices come through 510(k), the route for products similar to ones on the market: 329 of the 335 in 2025. De Novo grants, the route for new kinds of device, have never passed 9 in a year. Expect most future growth to come from new versions of familiar tools, not new types of product.

FDA AI Medical Devices by Specialty

Each device on the list is assigned a lead review panel, which works as a stand-in for the medical specialty it serves. Grouping by panel shows where AI has reached clinical products and where it has barely started.

• Radiology devices make up 1,230 of the 1,614 authorizations (76.2%). Their share since January 2025 is 75.4%, close to the all-time share.
• Cardiovascular is second at 154 devices (9.5%), followed by neurology at 73 (4.5%).
• Orthopedic devices are the fastest-growing small group: 6 of the 8 on the list were authorized in 2025 or 2026.
• Buyers outside radiology still have few authorized AI tools to choose from. Radiology holds 76.2% of all authorizations, against 9.5% for cardiovascular. If current shares hold, radiology will keep getting most new products.

Companies With the Most FDA AI Device Approvals

A small group of imaging manufacturers holds a large share of AI device authorizations. Subsidiaries and spelling variants on the FDA list are grouped under their parent company here.

• The five largest imaging manufacturers (GE HealthCare, Siemens Healthineers, Philips, United Imaging Healthcare and Canon Medical Systems) hold 343 authorizations, 21.3% of the list.
• Aidoc, which sells AI triage software for radiology, has 35 authorizations, the most of any company that does not make imaging hardware.
• Hospitals that buy scanners from the big imaging makers have an easy path to AI add-ons. Those five makers hold 343 authorizations, 21.3% of the list. Independent vendors such as Aidoc, with 35, are likely to compete on breadth of conditions covered.

FDA 510(k) Review Times for AI Devices

Review time is the number of days between the date the FDA received a 510(k) submission and its decision date. Both dates come from the FDA's 510(k) database, and the figures cover the 1,553 AI devices cleared through 510(k).

• The median AI device 510(k) took 136 days from receipt to clearance across all years, and 142 days for decisions from January 2025 to June 2026.
• Under its current user fee agreement, the FDA's goal is a 510(k) decision within 90 FDA days for 95% of submissions. The days counted here are calendar days and include time the clock was paused while the company answered FDA questions.
• For AI device companies, clearance takes longer than the FDA goal suggests. The median 510(k) took 142 calendar days from January 2025 to June 2026. Companies should plan for four to five months, or less through the Special 510(k) route for updates.

Where FDA AI Device Makers Are Based

The 510(k) database records the applicant's country, which shows how much of the AI device market comes from outside the US.

• US applicants hold 710 of the 1,553 AI device 510(k) clearances (45.7%), and 210 of the 505 cleared from January 2025 to June 2026 (41.6%).
• Israel, home to Aidoc and several other radiology AI companies, is second with 121 clearances.
• US hospitals are buying from a more international set of AI vendors. US applicants' share of clearances fell to 41.6% since January 2025, and France passed Israel in that period. That share of foreign vendors is likely to keep growing.

What FDA AI Medical Device Approval Trends Mean for 2027 and Beyond

Most AI tools the FDA has cleared are for medical imaging, and most came through the faster route for devices similar to ones on the market. That means hospitals have plenty of choice in imaging AI and far less in other areas of care.

A steadier pace of new products gives hospitals time to compare tools within a category before buying. A growing share of those tools will come from companies based outside the US, so contracts, support and service will more often involve vendors abroad. Growth outside radiology will depend on whether the FDA's process works as well for tools that do not fit the existing device categories.

So if your hospital is buying AI tools, check each product's FDA clearance and what it was cleared to do, because many tools in the same category were cleared for different uses.