healthcare
FDA AI-Enabled Medical Device Authorizations (Trailing 12 Months)
Latest
340authorizations, trailing 12 months
Methodology
## What this measures
The number of AI-enabled medical devices the FDA authorized in the 12 months to each month-end, from January 2015 to the last full month the list covers.
The FDA's list of AI-enabled medical devices: devices that use artificial intelligence (most of them machine-learning software that reads images or signals) and that the FDA authorized for marketing in the United States through 510(k) clearance, De Novo classification or premarket approval.
## Why it matters
It is the plainest count of AI reaching clinical use under regulation: each entry is a product a hospital or clinic may legally buy and use on patients.
## How to read it
The rolling count smooths the lumpy monthly decisions. The newest months can rise later, when the FDA adds devices to the list in its next update.
## What it does not say
The list holds the devices the FDA has identified from public summaries; the FDA says it is not a complete list, and it adds devices in batches a few times a year, so recent months fill in late. It counts authorizations, not devices in use or sold, and says nothing about how well a device works. Each 510(k) clearance counts once, so a new version of an existing product can count again.
## Source
U.S. Food and Drug Administration, "Artificial Intelligence-Enabled Medical Devices" list (fda.gov/medical-devices/artificial-intelligence-enabled-medical-devices), read from the CSV the list page links. Palanor counts devices by Date of Final Decision. FDA website policies: "Unless otherwise noted, the contents of the FDA website (www.fda.gov) — both text and graphics — are not copyrighted. They are in the public domain and may be republished, reprinted and otherwise used freely by anyone without the need to obtain permission from FDA." Read weekly by /api/cron/fda-ai-devices-sync.
Read this signal inside the lattice.
Palanor weighs every signal against the world your organization is watching. Methodology is public; integration is the platform.
See it in Palanor →