Pro se litigant Gregory Hardy filed a motion to consolidate cases one month after being warned about AI-hallucinated citations — and the new motion contained fresh fake quotations from real federal cases. The court struck the motion, noting that the Sixth Circuit had just confirmed even a single fak
A US court warned plaintiff’s counsel in Marino v. Marano after AI-generated citations for non-existent cases appeared in a civil brief filed March 25, 2026. The court ordered counsel to show cause why sanctions should not be imposed, reinforcing that attorney certification obligations apply fully t
Alabama federal judge publicly reprimanded attorney Edward May for submitting two fabricated quotations generated by ChatGPT and OpenCase AI. May admitted using AI in a multi-step drafting process, arriving late to his own show-cause hearing, and filing a typographically sloppy response — all of whi
Texas’s Seventh Court of Appeals upheld $14,271.25 in attorney’s fees against a pro se plaintiff whose AI-generated brief cited non-existent cases and contained fabricated quotations from real cases, with the trial court having already warned her that AI use does not excuse misrepresentations.
A Western District of Oklahoma court denied a plaintiff’s motion in limine after finding that cited Tenth Circuit cases did not support the propositions for which they were cited — a pattern consistent with AI-generated mischaracterization. The court warned that such mischaracterizations, whether fr
Pro se plaintiffs in a cryptocurrency credit line dispute cited a repealed Indiana statute and a nonexistent Seventh Circuit case — prompting a judicial warning that AI may have assisted the filings. No sanctions, but the court explicitly invoked the Seventh Circuit’s AI accuracy warning.
Texas’s Fifth District Court of Appeals affirmed a foreclosure judgment and separately identified that a citation in the appellants’ reply brief appeared to be a hallucinated case from this court that could not be located after extensive research.
Gregory Hardy filed the same AI-hallucinated motion to consolidate in Hardy v. Jones as he had in his other cases, repeating fake quotations from Berndt v. Tennessee and United States v. Frazier. The court struck the motion and issued a final warning: the next instance of fabricated citations or imp
A New York federal court sanctioned attorneys Levidow, Levidow & Oberman after they submitted a ChatGPT-generated brief citing six entirely fabricated cases, refused to acknowledge the errors when challenged, and then doubled down by filing fraudulent “confirmation” letters. The $5,000 sanction and
Ohio’s Eleventh District Court of Appeals sanctioned Attorney William Norman with a $2,000 fine, disciplinary referral, and a two-year court-notification requirement after his paralegal used ChatGPT to fabricate prosecutor quotes in a murder appeal, exposing a systemic failure in AI oversight.