legaltech.fyi Beta

Tracker

Sanctions watch

Every incident where a lawyer got caught filing AI-fabricated citations. The Hall of Shame.

Embed this
858
Incidents tracked
265
Incidents in 2026
$2,417,608
In fines, where reported
214 of 858 report an amount

The curve

Incidents per month

Last 18 months of published incidents, by the date of the court's order.

Apr 2025 Sep 2026

Charting 450 of 858 incidents; the rest are undated or fall outside this window.

Blamed by name

ChatGPT 151 Claude 9 CoCounsel 8 Copilot 8 Gemini 8 Lexis+ AI 6 AI-assisted legal software 4 Westlaw Precision 4 ChatGPT-4o 2 ChatOn 2 Eve 2 Google AI 2 LEAP 2 Perplexity AI 2 AI Overviews 1 Archie 1 Centient AI 1 ChatGPT-4 1 ClearBrief 1 CoCounsel, Westlaw Precision, Google Gem 1 Descrybe.ai 1 First Drafts 1 GPT-3.5 1 Ghostwriter Legal 1 Google Scholar 1 Grammarly 1 Grok 1 Irys 1 Legal Genius 1 LegalAI 1 MX2.law 1 OpenCase 1 Strongsuit 1 vLex Fastcase 1

Counts only incidents where the filing court or the reporting named the tool.

Hall of Shame

All Fined 251 Caught, not sanctioned 199 Warned 125 Referred to the bar 101 Reprimanded 86 Outcome unclear 56 Filing dismissed 37 Suspension 2 Upcoming sanction 1
Fined
A self-represented defendant filed submissions with AI-generated non-existent cases, and the court allowed correction but ordered costs.
Halton (Regional Municipality) v. Rewa et al. · Mr. Rewa · Ontario Superior Court
“one also does not need to be a lawyer to read through a case to verify if it stands for the suggested proposition.”
torkin.com ↗
The file →
  • Halton (Regional Municipality) v. Rewa et al. involved Mr. Rewa, a self-represented defendant.
  • He relied on AI-generated, non-existent cases in his court submissions.
  • The court allowed him to correct his motion but ordered costs.
  • The court was the Ontario Superior Court in Ontario.
  • The article says the judge warned AI-hallucinated case law would not be tolerated and quoted the court on verifying authorities.
Warned ChatGPT
An immigration barrister cited entirely fictitious or irrelevant cases in appeal grounds and the Upper Tribunal judge said he had likely used generative AI and tried to hide it.
Chowdhury Rahman · Chowdhury Rahman · Upper Tribunal
“The most obvious explanation is … that the grounds of appeal were drafted in whole or in part by generative artificial intelligence such as ChatGPT.”
theguardian.com ↗
The file →
  • Chowdhury Rahman was the immigration barrister in the Upper Tribunal case.
  • He cited cases that were described as entirely fictitious or wholly irrelevant.
  • The judge said Rahman likely used generative AI and considered reporting him to the Bar Standards Board.
  • Court: Upper Tribunal in the UK; the ruling was published on Tuesday.
  • Judge Mark Blundell ruled in September and said the submissions wasted the tribunal’s time.
Reprimanded ChatGPT
Two outside attorneys cited cases that ChatGPT hallucinated in an unrelated lawsuit and were publicly reprimanded and referred to licensing authorities.
Baltimore Central Booking & Intake Center conditions case · Matthew Reeves and William Lunsford · U.S. District Court · MD
thebanner.com ↗
The file →
  • Matthew Reeves and William Lunsford were the attorneys.
  • They cited cases that ChatGPT hallucinated in an unrelated lawsuit.
  • Judge Anna M. Manasco publicly reprimanded them and referred them to bar authorities.
  • U.S. District Court in Alabama.
  • The article says this happened this year; the page names ChatGPT and the referral.
Fined $10K ChatGPT
A California attorney was fined for filing a state court appeal with fake quotations generated by ChatGPT.
California attorney sanctioned for fake quotations generated by ChatGPT · CA
cybernews.com ↗
The file →
  • A California attorney was sanctioned for an AI-generated filing.
  • The attorney used fake quotations generated by ChatGPT in a state court appeal.
  • The court imposed a $10,000 fine.
  • The matter was in California state court.
  • The page says the fine was issued in September and calls it historic.
Caught, not sanctioned
Attorney Michael Fourte filed a brief using AI citations and quotes, and the court criticized him for relying on unvetted AI.
Michael Fourte · Michael Fourte · New York Supreme Court · NY
“Counsel relied upon unvetted AI – in his telling, via inadequately supervised colleagues – to defend his use of unvetted AI”
cybernews.com ↗
The file →
  • Michael Fourte defended himself in New York Supreme Court.
  • He submitted a brief using AI citations and quotes written with a chatbot.
  • The page does not report a formal sanction or amount for this incident.
  • The court was the New York Supreme Court in New York.
  • Judge Joel Coehn wrote the quoted criticism about unvetted AI.
Warned ChatGPT
A U.S. lawyer used ChatGPT to draft filings and cited fake cases; a federal judge responded with a standing order on AI use.
ChatGPT references nonexistent legal cases · A U.S. lawyer · federal judge
evidentlyai.com ↗
The file →
  • A U.S. lawyer in the ChatGPT case
  • He used ChatGPT and ended up citing entirely fake legal cases
  • The judge issued a standing order, not a monetary penalty
  • Federal court; the page gives no judge name or location
  • The page says the lawyer did not realize ChatGPT was a generative language tool
Caught, not sanctioned ChatGPT
Michael Cohen gave his lawyer AI-generated citations, and fake cases ended up in a real court filing.
Michael Cohen matter · Michael Cohen · U.S. District Court for the Southern District of New York · NY
“the attorneys’ actions were “embarrassing and certainly negligent.””
medium.com ↗
The file →
  • Michael Cohen and his lawyer were involved.
  • AI-generated citations were given to the lawyer and filed in court.
  • The page says they got sanctioned, but gives no order details.
  • No court, judge, or location is named on the page.
  • ChatGPT is identified as the tool used.
Caught, not sanctioned
A San Diego judge denied sanctions but condemned Tyson & Mendes for miscited and non-existent authority, factual misstatements, and unverified AI-sourced assertions.
Tyson & Mendes · Tyson & Mendes · San Diego court · CA
dailyjournal.com ↗
The file →
  • Tyson & Mendes was the firm criticized by the San Diego judge.
  • The filing contained miscited and non-existent authority and AI-sourced assertions.
  • The court denied sanctions despite condemning the conduct.
  • The court was in San Diego, California.
  • The article does not name an AI tool or quote the judge in the excerpt provided.
Filing dismissed
A self-represented appellant used generative AI to draft material with several fabricated case citations, and the court dismissed the appeal.
Helmod & Mariya (No 2) · Federal Circuit and Family Court of Australia (Division 1) Appellate Jurisdiction
shorestonelegal.com.au ↗
The file →
  • Helmod & Mariya (No 2) involved a self-represented appellant husband.
  • He used generative AI to prepare written material and filed several cases as authorities.
  • The court dismissed the appeal; no monetary sanction is stated.
  • Federal Circuit and Family Court of Australia (Division 1) Appellate Jurisdiction.
  • 2025; the court said the cases either could not be located or were not authority for the propositions.

Showing 151–160 of 858.

Get Sanctions Watch in your inbox
New AI-citation sanctions, the morning after they hit the docket.