Jarrus et al. v. Governor of Michigan et al.
E.D. Michigan · Dec 2, 2025
What happened in this matter?
The court sanctioned pro se plaintiffs $600 for violating Rule 11 by submitting filings containing three misrepresented, AI-generated case citations. The court determined that the plaintiffs failed to verify the accuracy of content produced by ChatGPT Plus, despite explicit prior warnings from the court regarding the risks of using generative AI tools in legal drafting.
- Why the court cared
- Misrepresented: Case Law | Plaintiffs cited Krupski for the proposition that relation-back amendment is mandatory; court found Krupski does not establish that mandatory rule and the plaintiffs overstated the case's holding. || Misrepresented: Case Law | Plaintiffs cited Heller to claim temporary loss of Second Amendment rights is irreparable harm; court held Heller does not state that and plaintiffs improperly extended Elrod (a First Amendment case) to the Second Amendment. || Outdated Advice: Repealed Law | Plaintiffs relied on Oppenheimer for scope of discovery under Rule 26; court noted Oppenheimer has been superseded by statute and reliance on it was misplaced.
- Why it matters now
- This record documents a Pro Se Litigant filing issue in E.D. Michigan, with the listed outcome: Monetary sanctions.
Why this matter is tracked
Misrepresented: Case Law | Plaintiffs cited Krupski for the proposition that relation-back amendment is mandatory; court found Krupski does not establish that mandatory rule and the plaintiffs overstated the case's holding. Outcome: Monetary sanctions. "[T]he fact that Plaintiffs … did not "fabricate cases or cite nonexistent decisions" is of no help. When a case cite is "real," an attorney, or for that matter a judge, might see a case they recognize and assume the quote or holding has been accurately represented. That problem is illustrated here; although Chat GPT generated "holdings" that looked like they could plausibly have appeared in the cited cases, in fact it overstated their holdings to a significant degree. And while a litigant might get away with similar overstatements because they could, perhaps, reason their way to showing how a case's stated holding might extend to novel situations, an LLM does not reason in the way a litigant must.To put it in a slightly different way, LLMs do not perform the metacognitive processes that are necessary to comply with Rule 11. LLMs are tools that "emulate the communicative function of language, not the separate and distinct cognitive process of thinking and reasoning." When an LLM overstates a holding of a case, it is not because it made a mistake when logically working through how that case might represent a "nonfrivolous argument for extending, modifying, or reversing existing law or for establishing new law;" it is just piecing together a plausible-looking sentence—one whose content may or may not be true."Court later declined to reconsider this decision (see here).
This record documents a Pro Se Litigant filing issue in E.D. Michigan, with the listed outcome: Monetary sanctions.
Record details
What the record establishes about AI use
Plaintiff Michael Jarrus identified the tool as ChatGPT Plus.
Opinion and Order on Order to Show Cause and Imposing Sanctions for Use of ChatGPT.
Correction behavior is not separately verified in the current record.
Tracked discrepancy record
3 citation, quotation, or authority issues are recorded in the source dataset.
- Misrepresented: Case Law | Plaintiffs cited Krupski for the proposition that relation-back amendment is mandatory; court found Krupski does not establish that mandatory rule and the plaintiffs overstated the case's holding.
- Misrepresented: Case Law | Plaintiffs cited Heller to claim temporary loss of Second Amendment rights is irreparable harm; court held Heller does not state that and plaintiffs improperly extended Elrod (a First Amendment case) to the Second Amendment.
- Outdated Advice: Repealed Law | Plaintiffs relied on Oppenheimer for scope of discovery under Rule 26; court noted Oppenheimer has been superseded by statute and reliance on it was misplaced.
Questions this record answers
- What happened in Jarrus et al. v. Governor of Michigan et al.?
- The court sanctioned pro se plaintiffs $600 for violating Rule 11 by submitting filings containing three misrepresented, AI-generated case citations. The court determined that the plaintiffs failed to verify the accuracy of content produced by ChatGPT Plus, despite explicit prior warnings from the court regarding the risks of using generative AI tools in legal drafting.
- Why does Jarrus et al. v. Governor of Michigan et al. matter for legal AI risk?
- This record documents a Pro Se Litigant filing issue in E.D. Michigan, with the listed outcome: Monetary sanctions.
- What does the public record establish about Jarrus et al. v. Governor of Michigan et al.?
- Retained unchanged after audit: the existing summary is source-specific, non-placeholder, and supported by the recorded source link. Read the underlying source before relying on it for legal advice or a filing.
- Which source supports this Jarrus et al. v. Governor of Michigan et al. summary?
- The recorded source is Damien Charlotin case archive. It is classified as publisher document archive; review the linked material and subsequent docket history before relying on this summary.
Related matters
Related by court, jurisdiction, tool, or classified failure pattern. Similarity does not imply the same facts or outcome.