legaltech.fyi Beta

Tracker

Sanctions watch

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

Embed this
857
Incidents tracked
266
Incidents in 2026
$2,397,608
In fines, where reported
212 of 857 report an amount

The curve

Incidents per month

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

May 2025 Oct 2026

Charting 447 of 857 incidents; the rest are undated or fall outside this window.

Blamed by name

ChatGPT 147 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 249 Caught, not sanctioned 198 Warned 124 Referred to the bar 101 Reprimanded 86 Outcome unclear 57 Filing dismissed 39 Suspension 2 Upcoming sanction 1
Filing dismissed 2025-10-01
ASBCA struck a law firm's reply brief after finding over 70% of the citations were inaccurate and included fabricated citations.
ASBCA reply brief sanction (October 2025) · Armed Services Board of Contract Appeals
“The ASBCA called the safeguards “woefully insufficient.””
burr.com ↗
The file →
  • An unnamed law firm was sanctioned by the ASBCA for Gen-AI misuse.
  • Its reply brief used AI technology and more than 70% of the citations were inaccurate.
  • ASBCA struck the reply brief as the sanction.
  • Armed Services Board of Contract Appeals, federal.
  • October 2025; government counsel and the Board found even more fabrications than the government had flagged.
Caught, not sanctioned ChatGPT 2025-09-30
Jisuh Lee filed a factum with ChatGPT-generated fake cases and later admitted she misled the court about who prepared it; the court treated it as criminal contempt.
Ko v. Li · Jisuh Lee · Ontario Superior Court of Justice
““Delivering a factum containing non-existent legal submissions creates a serious risk of a miscarriage of justice.””
benmor.com ↗
The file →
  • Ko v. Li involved lawyer Jisuh Lee and a factum with fabricated authorities.
  • The article says she used ChatGPT and failed to verify the citations.
  • The court treated the matter as criminal contempt, with no sanction yet imposed on this page.
  • Ontario Superior Court of Justice in Ontario, Canada.
  • On 2025-09-30, Lee admitted she alone used ChatGPT and had lied about staff involvement.
Filing dismissed ChatGPT 2025-09-30
A relator disclosed an expert report with generative-AI hallucinations, and the federal judge dismissed the FCA suit after related sanctions motions were filed.
U.S. ex rel. Khoury v. Intermountain Healthcare Inc et al. · Gary Giampetruzzi, Jessica R. Montes and Vanna Mavromatis · U.S. District Court for the District of Utah · UT
paulhastings.com ↗
The file →
  • U.S. ex rel. Khoury v. Intermountain Healthcare Inc et al. involved relator-side counsel and an expert report.
  • The report contained generative AI hallucinations, including bogus deposition testimony and fictitious quotes from government manuals.
  • On 2025-09-30, the federal judge granted the United States’ motion to intervene and dismiss the case.
  • The case was dismissed with prejudice to the relator and without prejudice to the United States.
  • District of Utah; the article does not name the AI tool, but says the expert used a generative AI tool and the parties had motion practice over sanctions.
Warned Copilot 2025-09-29
A lawyer filed fabricated and misrepresented case law using Microsoft Copilot, and the tribunal issued a show-cause order.
ANPV & SAPV v Secretary of State for the Home Department · Upper Tribunal
damiencharlotin.com ↗
The file →
  • ANPV & SAPV v Secretary of State for the Home Department involved a lawyer filing.
  • Microsoft Copilot was the named AI tool.
  • The filing contained fabricated and misrepresented case law, and the tribunal issued a show-cause order.
  • Upper Tribunal, UK.
  • 29 September 2025; the page lists a show-cause order.
Fined $620 2025-09-29
Mr Lalwani cited a fictitious AI-generated authority in written submissions and was ordered to pay S$800 in personal costs.
Tajudin bin Gulam Rasul and another v Suriaya bte Haja Mohideen · Mr Lalwani Anil Mangan · General Division of the High Court of the Republic of Singapore
“It follows that the citation of a fictitious AI-generated authority to the court, whether intentionally or otherwise, is wholly impermissible.”
elitigation.sg ↗
The file →
  • Tajudin bin Gulam Rasul and another v Suriaya bte Haja Mohideen; Counsel for the Claimants
  • Written submissions cited a fictitious authority produced by a generative AI tool
  • Court ordered CC to pay costs personally to the Defendant
  • General Division of the High Court of Singapore
  • 29 September 2025; the judge said fictitious AI-generated authority is wholly impermissible
Fined $18K 2025-09-26
Counsel filed a factum with hallucinated case authorities, and the court later ordered costs paid personally by appellant's lead counsel.
Reddy v. Saroya · Christopher Souster of Nimmons Law Office · Court of Appeal of Alberta · AB
“When used without safeguards, large language models frequently introduce confusion and delay into proceedings, and worse, constitute an abuse of process that may potentially bring the administration of justice into disrepute.”
canadianlawyermag.com ↗
The file →
  • Reddy v Saroya involved the appellant's lead counsel.
  • The factum contained fabricated case authorities and seven cited cases could not be found.
  • The court later ordered $17,550 plus GST payable personally by counsel.
  • Court of Appeal of Alberta, Alberta.
  • The page links the issue to a large language model but does not name a specific tool.
Caught, not sanctioned 2025-09-25
A petition was withdrawn after opposing counsel exposed fabricated citations that invented paragraphs from Raj Narain v. Indira Nehru Gandhi.
Greenopolis Welfare Association (GWA) v. Narender Singh et al. · Delhi High Court · DL

Cases that don't exist

  • Raj Narain v. Indira Nehru Gandhi
medianama.com ↗
The file →
  • Greenopolis Welfare Association v. Narender Singh was a flat-possession dispute.
  • The petition quoted invented paragraphs from a landmark judgment.
  • The petition was withdrawn after the fabricated citations were exposed.
  • Delhi High Court, Delhi.
  • September 2025; it quoted paragraphs 73 and 74 from a 27-paragraph judgment.
Reprimanded 2025-09-25
A plaintiff's lawyer quoted non-existent BGH decisions; the Frankfurt court condemned inventing citations and using chatbot-suggested sources.
LG Frankfurt, 2-13 S 56/24 · LG Frankfurt
“it is “among the fundamental obligations of the legal profession neither to invent citations nor to include unverified sources suggested by a chatbot in a court filing.””
posteo.de ↗
The file →
  • LG Frankfurt, 2-13 S 56/24 involving a German lawyer.
  • Chatbot produced three verbatim BGH quotes that were complete fabrications.
  • Issued a formal judicial rebuke.
  • Landgericht Frankfurt (Germany).
  • Dated September 25, 2025; court labeled the AI-generated authorities complete fakes.
Filing dismissed ChatGPT 2025-09-25
Oready, LLC filed protest briefs with fabricated, mismatched, and misquoted GAO citations, and GAO dismissed the protests as an abuse of process.
Oready, LLC · Government Accountability Office
“Our decision here is not based on the use of AI as a method of research; it is based on the protester’s repeated reliance on non-existent citations or decisions without verifying their correctness and applicability to the arguments being presented.”
centrelawgroup.com ↗
The file →
  • Oready, LLC was dismissed after filing AI-tainted protest briefs.
  • The filings included fabricated GAO decisions, mismatched citations, and misquoted holdings.
  • GAO dismissed the protests as an abuse of process.
  • Government Accountability Office; federal procurement protest forum.
  • Sept. 25, 2025; GAO said the issue was unverified non-existent citations, not AI use itself.

Showing 701–710 of 857.

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