FDA cleared a record number of AI/ML devices in 2025 — what research teams should learn

What the numbers say (cite trackers carefully)

Independent analyses of the FDA’s public AI/ML-enabled device list report a record ~295 clearances in 2025, with cumulative authorizations on the order of ~1,450 by end of 2025 (exact totals move as the FDA updates the list). Radiology / imaging continues to dominate clearances.

Useful roundups: Innolitics 2025 year in review · FDA’s AI-enabled medical device list.

Separately, FDA has issued draft guidance materials on credibility assessment for AI used in regulatory decision-making and AI-enabled device development—developers of devices must follow FDA pathways; research frameworks do not substitute for clearance.

What developers of devices typically need to demonstrate

• Clear intended use and risk characterization
• Validation appropriate to the claim (including diverse populations where relevant)
• Transparency / labeling commensurate with risk
• Lifecycle monitoring plans (including predetermined change control where applicable)

Important: The EHR Risk Framework is not an FDA-cleared device and is not a shortcut to 510(k)/De Novo. It is research/education software.

How the EHR Risk Framework helps (pre-device research culture)

• Leakage-aware evaluation habits before anyone talks about “performance”
• Calibration metrics (Brier / ECE) for probability quality
• SHAP for inspectable drivers in teaching demos
• Reproducible Docker + report ZIP workflows

Sources

• Innolitics. 2025 Year in Review: AI/ML Medical Device 510(k) Clearances. innolitics.com
• U.S. FDA. AI-Enabled Medical Devices (public list, scope notes, and authorization records). fda.gov

Try the workbench

How it works A–Z Live demo Quickstart GitHub

Free demo server — it may be slow. Check it with a small amount of data. For larger workloads or freer experimentation, run locally or on your own server. Open live demo

Feedback: support@larucare.com · Cite: DOI & CITATION.cff