Lawyers Sanctioned for AI: A Tracker of Fake-Citation Cases
Which lawyers and firms have been sanctioned for AI-fabricated citations, what happened in each case, and the verification steps that keep you out of the next one.
Lawyers sanctioned for AI hallucinations now appear in court dockets across the country. Since 2023, judges have fined attorneys, revoked their status, and referred them for discipline after they filed briefs with fake case citations from AI tools.
An AI hallucination in a legal filing is a citation that looks real but does not exist. The case name, the court, and the quote can all be invented by a chatbot that was never checking a legal database.
This page tracks the real, documented cases. It explains why AI invents fake citations, what courts now require, and the verification steps that keep your firm out of the next sanctions order.
Why AI Invents Fake Case Citations
AI invents fake case citations because a large language model predicts plausible text, not verified law. Tools like ChatGPT and Gemini generate the words most likely to follow your prompt, and a realistic-looking citation is exactly that kind of text.
These models are not legal databases. Unless a tool is connected to a real source like LexisNexis or Westlaw, it has no way to confirm a case exists. It simply produces a caption, a reporter number, and a quote that fit the pattern.
That is why hallucinated citations are so dangerous. They carry the right format, a confident tone, and a made-up holding that supports your argument. Nothing on the page warns you the case is fiction.
Worried a hallucinated citation could slip into a filing? We help firms put AI verification guardrails and a clear policy in place so it never happens.
Book a ConsultationAI Hallucination Sanctions: Case Tracker
The cases below are verified sanctions and discipline actions tied to AI-fabricated citations. Each row links to a court opinion or established news report so you can read the source yourself.
| Case | Court | Year | What went wrong | Outcome |
|---|---|---|---|---|
| Mata v. Avianca | S.D.N.Y. | 2023 | Two lawyers filed a brief with fake cases invented by ChatGPT, then defended them when challenged | Judge P. Kevin Castel imposed a $5,000 sanction under Rule 11 for bad faith |
| Park v. Kim | 2nd Cir. | 2024 | A lawyer cited a nonexistent case generated by ChatGPT in a reply brief | The attorney was referred to the court's grievance panel for discipline |
| Wadsworth v. Walmart | D. Wyo. | 2025 | Morgan & Morgan lawyers filed a motion with eight fake cases from an in-house AI tool | The lead lawyer was fined $3,000 and lost pro hac vice status; two others were fined $1,000 each |
| Coomer v. Lindell | D. Colo. | 2025 | MyPillow defense lawyers filed a brief with nearly 30 errors and hallucinated cases | Judge Nina Wang fined two lawyers $3,000 each |
| Johnson v. Dunn | N.D. Ala. | 2025 | Three Butler Snow lawyers filed motions with five fake cases from ChatGPT | Judge Anna Manasco removed them from the case, reprimanded them, and referred them to the state bar |
| Cohen matter | S.D.N.Y. | 2024 | Michael Cohen gave his lawyer fake cases from Google Bard (now Gemini), which were then filed | Judge Jesse Furman declined to sanction, finding no bad faith by Cohen |
The last row is a useful contrast. Judges weigh intent, so a lawyer who owns the mistake early is treated very differently from one who defends a fake case after being caught.
What Courts Now Require
Many courts now require lawyers to disclose AI use or certify that they checked every citation by hand. These rules arrived fast after the first fake-citation scandals and keep spreading through standing orders and bar guidance.
The first example set the pattern. In May 2023, Judge Brantley Starr of the Northern District of Texas required every attorney to certify that either no filing was drafted by AI, or that a human checked any AI-drafted language for accuracy.
Other courts across the country modeled their orders on that framework. Even where no special order exists, the duty is not new. Rule 11 already requires that every filing be grounded in real fact and law, which is enough to sanction a fake citation.
How Lawyers Can Avoid AI Sanctions
You avoid AI sanctions by verifying every citation against the official record before you file. AI can speed up drafting, but the lawyer who signs the brief is responsible for every case in it.
The checklist below is the core of a safe workflow. It works whether you use a general chatbot or a legal-specific tool that links to sources.
- Pull every case yourself in the official reporter, LexisNexis, or Westlaw. If you cannot find it there, do not cite it.
- Read the actual opinion and check that each quote and holding matches the source. AI often invents the quote too.
- Never file AI output unread. Treat a draft from a chatbot as an unverified starting point, not a finished brief.
- Disclose AI use when your judge or local rules require it, and keep your certification current.
- Prefer legal tools that link each citation back to a real source, but still confirm every case before filing.
The Bottom Line
Lawyers sanctioned for AI hallucinations share one thing in common: nobody checked the citations before filing. The tools invented plausible cases, and the humans trusted them.
The fix is not to ban AI. It is to build a verification step that no filing can skip, so a hallucinated citation is caught long before a judge sees it.
A short policy and a firm-wide checklist are usually enough. The cost of setting them up is trivial next to a public sanctions order and a bar referral.
Frequently Asked Questions
- Yes. Since 2023, judges have fined lawyers, revoked their court status, and referred them for discipline for filing briefs with fake AI-generated citations. Documented cases include Mata v. Avianca, Wadsworth v. Walmart, Coomer v. Lindell, and the Butler Snow matter in Alabama.
- Mata v. Avianca is the landmark case. In 2023, Judge P. Kevin Castel of the Southern District of New York sanctioned two lawyers $5,000 after they filed a brief with fake cases invented by ChatGPT and then defended the citations when challenged.
- You are not disbarred for using ChatGPT itself, but filing fake citations can trigger discipline that ranges up to disbarment. Courts have already imposed fines, removed lawyers from cases, and referred attorneys to state bars for AI-fabricated citations. The offense is filing unverified work, not the tool.
- Lawyers verify AI citations by pulling each case themselves in an official reporter or a database like LexisNexis or Westlaw. They confirm the case exists, then read the opinion to check that every quote and holding is accurate before filing.
- AI makes up fake cases because a language model predicts plausible text rather than retrieving verified law. Unless it is connected to a real legal database, it generates a citation that fits the pattern, complete with a fake court, year, and quote that look real.
- Legal-specific tools that link to real sources reduce hallucinations but do not eliminate the lawyer's duty to verify. Even tools built on LexisNexis or Westlaw data can misstate a holding, so you still confirm every citation against the source before you file.
Put AI Guardrails in Place Before You File
We help law firms build the verification workflow and AI-use policy that keep a hallucinated citation from ever reaching a court. Book a workflow audit to see where your process needs a guardrail.
Book a Workflow Audit