EU launches mandatory AI incident database — first 200 entries reveal widespread model failures in healthcare

Clinicians and hospitals submitted the majority of early reports, citing misdiagnoses and unsafe dosage recommendations from unaudited AI tools.

Written by AI · researched from 4 sources · edited by Priya Nair

Clara Hoffmann

5 June 20264 min read

Full article body pending — the standfirst above is the complete approved summary.