Jimenez-Fogarty v. Fogarty

What sanction did the court impose?

The court found a Rule 11 violation because counsel relying on non-existent authority "made no inquiry, much less the reasonable inquiry required by Rule 11," and found that "Lindsay used an AI program that generated the fabricated citations." It found subjective bad faith, which the Second Circuit requires for a sua sponte Rule 11 sanction, and, accepting arguendo that a clear-and-convincing standard applies, found it met, based on the AI-generated citations, her failure to give the ordered drafting account, her minimizing statements, and her repeat filings. It sanctioned her $2,500, and ordered her by May 14, 2026 to pay, give her client a copy of the opinion, alert the judge in each pending case where she is attorney of record by a letter filed on the docket, and file a sworn statement of compliance. Those deadlines extend automatically to 7 days after disposition of any timely Rule 72 objections. It declined to award Hirshowitz fees. The docket shows an extension order on May 14, 2026 (ECF No. 195) and, on May 28, 2026, an "Objection to Report and Recommendations" with Lindsay's affirmation (ECF No. 196), with no ruling on it listed through July 1, 2026; neither document is in RECAP, so whether ECF No. 196 objects to this order is unconfirmed.

Why does Jimenez-Fogarty v. Fogarty matter for law firms using AI?

Judge Gorenstein’s opinion is built around a question counsel would not answer: where did the fabricated citations come from? The show-cause orders asked for “a complete and detailed description of the process of the drafting” and of “the role of any individual or computer system.” The court found that Lindsay’s sworn response offered “only airy generalities and conclusory statements,” and that her description of manually cross-checking citations could not be true, because such a check “would have instantly shown that each citation was fake or inaccurate.”

The court then reasoned by elimination. With no evidence that an established research platform generates fabricated citations, and with Lindsay denying that any other person helped draft the briefs, “there remain only two plausible explanations”: a person who fabricated them knowingly, or an AI system. The court concluded that an AI program generated them, while expressly declining to credit her later suggestion that it was a LexisNexis product. It then quoted Braica v. Frankowski for the proposition that presenting AI hallucinations as valid caselaw has repeatedly been found to be subjective bad faith in the Second Circuit, and added further grounds: her failure to comply with the drafting-description order, her minimizing statements, and her repeat filings after the show-cause orders.

The opinion distinguishes two district court decisions in the circuit that found no bad faith, United States v. Cohen and Hall v. Academy Charter School, on the ground that the lawyers there had some reason to trust the source of the citations and apologized or fully explained what happened. Firms documenting compliance may wish to consider that, in this case, the lack of a candid, specific account of how the citations were produced weighed as heavily as the citations themselves.