Wiki · Enforcement timeline
Catalog dashboard
Enforcement timeline
Chronologically-ordered enforcement actions linked to the instruments and topics tracked in the Policy Window catalog. Closed cases sort by their resolution year; ongoing cases sort by initiation. Each entry links to the catalog topics + the catalog instruments the case touches, and (where possible) to the best-available primary source for the action.
13 catalogued cases total · 1 match the current filter (topic = AI in Healthcare) clear all.
Filters
Cases (1)
- Settlement with remedy
- Open arrow = ongoing (no resolution year yet)
- Settlement with remedy
2024 · US
Texas AG v. Pieces Technologies (healthcare AI deceptive practices)
Enforcer: Texas Attorney General
Target: Pieces Technologies, Inc.
Violation alleged: Pieces marketed its generative-AI clinical-summary product to Texas hospitals with materially misleading accuracy claims (alleging severe-hallucination rates orders of magnitude lower than measured in deployment).
Lesson: First US state-attorney-general action against a generative-AI vendor under state UDAP/UDTPA authority. Settlement (Sep 2024) required: (1) clear disclosure of AI involvement to end-users, (2) accurate marketing-claim substantiation, (3) ongoing monitoring of model output for hallucination rates. Establishes that state AGs can reach AI deployment claims without an AI-specific state statute — through long-standing consumer-protection authority. Likely template for parallel actions by other state AGs.
Source: https://www.texasattorneygeneral.gov/news/releases/attorney-general-ken-paxton-secures-settlement-healthcare-technology-company-pieces-technologiesregulator landing
Editorial scope
This timeline reflects enforcement cases tracked in the Policy Window catalog. It is not exhaustive; coverage focuses on high-precedent matters relevant to the catalogued instruments (currently 13 cases across EU, US, UK, China, India, Italy, France). The catalog deliberately omits routine regulator letters and ongoing investigations whose materials are not public.
For the inclusion rubric (when a case enters the catalog, what level of source-defensibility is required, how jurisdictional balance is managed), see /wiki/methodology. Cases marked “ongoing” remain editorial-watch items; outcomes get backfilled as the public record settles.