Trackers · Living reference
AI Incident & Failure Tracker
What actually went wrong — documented production AI failures, liability rulings, and court sanctions, each with a primary source. The reference for anyone writing an AI policy who needs precedent, not vibes. Updated as cases land.
Last updated 2026-07-16 · 11 incidents
| Date | Subject | What happened | AI involved | Type | Source |
|---|---|---|---|---|---|
| twigg v. bsn sports (m.d. pa.) | Pennsylvania plaintiff's attorney | Chief Judge Matthew W. Brann fined the attorney $1,500 and suspended him from practicing in the Middle District of Pennsylvania for six months for a Rule 11 violation after filings contained AI-generated hallucinated citations in Twigg v. BSN Sports, Inc., No. 4:23-CV-00067-MWB. | Unspecified generative AI | Court sanction | Law.com / The Legal Intelligencerlaw.com |
| withers v. city of aberdeen (n.d. miss.) | Kathleen M. Wilson, Kathryn Y. Williams, Shauncey Hunter Ridgeway, Mark C. McClinton | Senior Judge Sharion Aycock disqualified all four attorneys from the case, revoked the two out-of-state lawyers' pro hac vice admissions and barred them from the Northern District of Mississippi for two years, and imposed $8,000 in combined fines ($2,500 Wilson, $3,500 Williams, $1,000 each for local counsel) after both sides filed AI-hallucinated citations in the Withers v. City of Aberdeen fee dispute (No. 24-cv-218). | First Drafts and an unnamed AI research tool | Court sanction | Bloomberg Lawnews.bloomberglaw.com |
| ibach v. stewart (ala. sup. ct.) | W. Perry Hall (Mobile solo practitioner) | Alabama Supreme Court dismissed the appeal as frivolous and ordered Hall to pay $17,200 in attorney's fees and costs, referred him to the Alabama State Bar for discipline, and barred him from filing anything else in the court unless co-signed by another attorney in good standing — after briefs contained fabricated AI-hallucinated citations in the Ibach v. Stewart family-trust dispute (No. SC-2025-0106). | Unspecified generative AI | Court sanction | Bloomberg Lawnews.bloomberglaw.com |
| Apr 2026 | Brigandi order (D. Or., Mag. J. Clarke) | Roughly $96,000 sanction (≈$110K with related penalties) for AI-fabricated content across three filings — 15 references to nonexistent cases and 8 fabricated quotations — one of the largest single-attorney AI-hallucination penalties in U.S. federal court to date. | Generative AI (unspecified) | Court sanction | ABA Journalabajournal.com |
| Feb 2026 | US v. Heppner (S.D.N.Y., J. Rakoff) | Court held that materials a party generated with a consumer-grade AI tool were NOT privileged — inputting client confidences into a public LLM with no confidentiality obligation is a third-party disclosure that waives attorney-client privilege. | Consumer/public LLMs | Liability ruling | Morgan Lewismorganlewis.com |
| Feb 2026 | Warner v. Gilbarco (E.D. Mich.) | Court held AI-assisted work product WAS protected and not discoverable — "generative AI programs are tools, not persons" — when used at counsel’s direction. | Generative AI (general) | Liability ruling | JD Supra / Morgan Lewisjdsupra.com |
| Oct 2025 | California attorney (Mostafavi) | Fined $10,000 — the largest California AI-fabrication penalty to date — after 21 of 23 case quotes in an appellate brief turned out to be fabricated by ChatGPT. | OpenAI — ChatGPT | Court sanction | The Daily Recordthedailyrecord.com |
| May 2025 | Butler Snow (N.D. Ala., J. Manasco) | Show-cause order after a partner used ChatGPT for fake citations; the 300+-lawyer firm called the lapse "inexcusable" and three lawyers were removed from the case. | OpenAI — ChatGPT | Court sanction | ABA Journalabajournal.com |
| Feb 2025 | Morgan & Morgan attorneys (D. Wyo., J. Rankin) | $5,000 sanction for 8 fabricated citations; lead lawyer removed from the case after using the firm’s in-house tool MX2.law without verifying the cases. | MX2.law (in-house) | Court sanction | LawNextlawnext.com |
| Feb 2024 | Air Canada | Held liable for negligent misrepresentation by its support chatbot (Moffatt v. Air Canada); the tribunal rejected the airline’s claim that the chatbot was a "separate legal entity". | In-house support chatbot | Liability ruling | American Bar Associationamericanbar.org |
| Jun 2023 | Mata v. Avianca (two NY attorneys) | The landmark case: lawyers sanctioned after submitting a brief full of non-existent cases hallucinated by ChatGPT — the first widely-cited AI legal-citation failure. | OpenAI — ChatGPT | Court sanction | Seyfarth Shawseyfarth.com |
How this is compiled
- Documented incidents only. Each row cites court/tribunal coverage, law-firm analysis, or Tier-1 reporting — not rumor or speculation.
- Practitioner focus. We track the failures that set precedent for enterprise AI policy: who was held liable, who was sanctioned, and why.
- Know an incident we’re missing? Tell us.
Informational only · not legal advice