AI Hallucination Cases

The court found Attorney Erin Rich submitted briefings that contained false or embellished factual statements drawn from medical records (e.g., misstating a speech-therapy note, misdating primary-care records, and asserting “numerous seizures” when the record showed a single generalized seizure). Rich admitted reliance on prior work product and that her firm occasionally uses AI to summarize records, but could not identify a specific tool. The court held the falsities violated Rule 11, found the conduct willful within the firm’s chain of command, and imposed non-monetary sanctions (CLE courses including ethics/AI topics and an order to review pending motions) to deter repetition.

The Court found that attorney Tyrone A. Blackburn repeatedly used quotation marks for language that does not appear in the cases he cited (including at least 17 instances) and defended that practice as paraphrase. Opposing counsel (Roc Nation) flagged these apparent AI-generated citation errors, and the Court independently verified the quotations. Citing prior orders and sanctions (including a $5,000 sanction in Jakes v. Youngblood), the Court concluded Blackburn's conduct evidenced a pattern of fabricating quotations, granted the motion to strike his filing, deemed the sanctions motion unopposed, and referred Blackburn to the Grievance Committee. The Court emphasized the obligation to accurately quote and cite authority and noted refusal to accept paraphrases presented as verbatim quotations.

The court determined Plaintiff's emergency motion for a temporary restraining order was apparently drafted using generative AI and contained bracketed placeholders and a fabricated case citation ("Corrigan v. City of Scottsdale, 720 F.3d 513, 520 (9th Cir. 2013)"). The court noted this raised Federal Rule of Civil Procedure 11 concerns but, for judicial efficiency, addressed the motion on the merits, denied the TRO, and dismissed the complaint under 28 U.S.C. § 1915(e)(2)(B). No professional sanctions or monetary penalties were imposed related to the fabricated citation.

Earlier decision noticing the issues here.

In an innovative approach, the court set the fine amount by reference to the annual cost of a verified legal AI tool (citing products such as Sof-IA by Tirant lo Blanch, GenIA-L by Lefebvre, Harvey AI, and others), reasoning that the use of such a tool would have eliminated or drastically reduced the risk of hallucination

The Supreme Court found that the NCLT (and by affirmation the NCLAT) relied on AI-generated, non-existent case law and misattributed paragraphs presented as precedents. The Court held such reliance amounts to a subversion of the rule of law and declared zero tolerance for citing unverified AI-generated precedents, set aside the impugned orders, restored the Section 7 petition, and directed the Bar Council of India to formulate guidance and disciplinary measures.

The City Employees alerted the Court that Petrini's response relied on AI which 'hallucinated' a case holding and altered facts from Petrini's earlier Case No. 178—specifically attributing a roof stop-work order and related holding to Case No. 178 when that order contains no such material. The Court accepted this observation, reiterated warnings about AI-generated 'legal fiction' (citing Ferris and Fletcher), admonished the pro se plaintiff and reminded litigants of duties of candor and Rule 11, but imposed no sanctions based on the record.