Mata Tracker · Loading…

Every AI-sanction case we’ve found. Annotated.

A searchable database of every U.S. case where attorneys, experts, or judges were sanctioned, referred, or reprimanded for AI-generated hallucinations in court filings. Each case gets a structured practitioner analysis: Model Rules breakdown, bar referral risk, malpractice exposure, and the specific workflow that would have prevented the failure.

Browse all cases Methodology

—

Published Cases

— indexed

—

Total Penalties

—

Bar Referrals

—

Lawyer Cases

—

Rules & Orders

How the tracker works.

Sourcing

Cases are initially identified through Damien Charlotin's AI Hallucination Cases database as a research index. Primary-source opinions are independently retrieved from CourtListener and the RECAP archive on Internet Archive. We index only U.S. cases involving sanctions, referrals, or formal reprimands.

Annotation

Each case receives an AI-drafted practitioner overlay generated by a Claude Skill from the actual judicial opinion text. Annotations are reviewed by Blake Beyel (Florida Bar) before they go live; the rule references and risk assessments are educational commentary, not legal advice as to any specific matter. The annotation covers Model Rules 1.1, 1.4, 1.6, 3.3, 5.1, 5.3, and 8.4, plus insurance exposure, bar referral risk, and recommended firm policy language. See the Disclaimer for full scope.

Editorial Standard

The AI does the research at scale. A human does the judgment. Every annotation is reviewed by Blake Beyel (Florida Bar, JD/MBA/MAcc) before publication. No annotation goes live without a manual review pass. Corrections at blake.beyel@beyel.com.