In the case of Jimenez-Fogarty v. Fogarty, a ruling issued on Wednesday by Magistrate Gabriel W. Gorenstein of the Southern District of New York addressed the conduct of a candidate who stood for the New York State Senate in 2024.
Tricia S. Lindsay, who represented the plaintiff Sai Malena Jimenez-Fogarty, submitted two responses to motions to dismiss that relied on a large number of cases that could not be found. In response, the Court ordered Lindsay to show cause why she should not be sanctioned for these misleading filings. After reviewing Lindsay’s reply to those orders, the Court determined that a sanction of $2,500.00 was appropriate ….
[T]wo briefs bore Lindsay’s signature, and each contained numerous citations that could not be corroborated. By “fabricated,” the Court does not refer to occasional typographical slips—for instance, references to cases that exist and support the proposition but cannot be located in the volume or page indicated (or database identifier) given. Likewise, the Court excludes citations that, while correct in other respects, identify the wrong case name. It also excludes instances where the cited case covers a similar topic but its holding is mischaracterized, even though such a citation could itself be considered misleading and perhaps worthy of sanctions.
Instead, the Court considers only citations to cases that cannot be located by name at all—or cases that were not found where Lindsay claimed they would be and that have no relation to the propositions for which they were cited—thus demonstrating that the citations were entirely fabricated. The following passages describe some of the fabricated citations [seven items described in the record -EV] ….
Lindsay’s answer when confronted about the fabricated citations did not follow the usual pattern seen in matters where courts require attorneys to explain the presence of such citations. The Court’s first Order to Show Cause directed Lindsay to provide “a complete and detailed description of the drafting process for the two memoranda of law.” While most attorneys supply such an explanation—often admitting reliance on artificial intelligence tools—Lindsay’s sworn response offers only vague generalities and conclusory statements. The response contains no coherent explanation for how the two memoranda of law came to include the fabricated citations. The most basic questions—most notably: what was the source of the fabricated citations?—are left unaddressed.
Thus, Lindsay asserts only that she adheres to a “structured and diligent process designed to ensure accuracy, thoroughness, and compliance with all applicable legal and ethical standards.” Curiously, she then discusses her “typical” process for drafting briefs without specifically attesting that this was the actual process followed in drafting the memoranda at issue. The implication is that her usual process was in fact used here; otherwise, there would be no reason to mention it. Lindsay explains that she “manually cross-checks” case names, docket numbers, and reporter citations against the primary sources in legal databases and verifies that “the authorities cited support the propositions for which they are referenced.” Yet, her insistence that these steps occurred in this case is plainly false, because a truly manual check would have detected that the citations were fake or inaccurate. Accordingly, while Lindsay acknowledges that eleven citations were incorrect, she offers no explanation of how these citations ended up in the memoranda of law.
At one point, Lindsay attributes the existence of citations to “nonexistent cases” to “typographical errors, misreading of secondary sources, or limitations in the search functionality of legal databases.” In addition to being conclusory, this explanation makes little sense, since Lindsay provides no demonstration of how such errors, misreadings, or search limitations produced the nonexistent cases cited. And in any event, she offers no explanation for why her supposed “manual” checks did not uncover the nonexistent citations.
She also emphasizes that she relied on “established legal research platforms.” Yet the record does not support that such platforms ever generated or housed fabricated authorities, nor that any such platform provided the fabricated material in the memoranda. Notably, in her initial response, Lindsay never identified which “legal research platforms” she used, nor did she specify which one allegedly supplied the fabricated citations.
Given the complete absence of evidence that any “established” platform could have produced these citations, the Court rejects this claim. Consequently, two plausible explanations remain for the source of the fabricated material: a person deliberately creating them knowing they were false, or an AI system. Lindsay does not claim any human assistance in drafting the memoranda. In her declaration she states:
I am solely responsible for the research, drafting, and review of the memoranda at issue. While I employ legal research databases and citation-management tools to assist my work, I do not delegate these core tasks to others. No other individual or computer system played a substantive role in the drafting process beyond the use of standard legal research and word-processing software.
By denying any human help, Lindsay thus seeks to persuade the Court to accept a proposition unsupported by any evidence: that an established legal research platform created the fabricated citations.
In her response to Hirshowitz’s filing, Lindsay first-time states that she routinely uses LexisNexis’s AI-enabled features, and then suggests that the fabricated citations “may well have been generated by the LexisNexis software during the research phase.” The response, however, provides no specifics about which LexisNexis AI features were used, how they could have supplied any of the false citations, or any example where LexisNexis’s AI tools produced fabricated entries.
{In her reply to the Court’s second OSC addressing Rule 11, Lindsay again failed to describe how the fabricated citations ended up in her brief. Instead, she claimed that the citation errors were not the result of any intent to mislead the Court or opposing counsel but rather “isolated and inadvertent mistakes made in good faith efforts to ensure accuracy.” She also states that she “relied on technology that is not immune to error” without naming the technology.}
From this, the Court concludes that Lindsay used an AI program to generate the fabricated citations and failed to verify that those citations were legitimate….
“[C]ourts in this circuit have repeatedly held that presenting AI-generated hallucinations as valid caselaw constitutes subjective bad faith.”
Lindsay’s bad faith is underscored by her failure to provide the “complete and detailed description of the drafting process” and the “detailed and complete description of the role of any individual or computer system involved” in the drafting, as required by the Court. Her evasive posture reflects a broader unwillingness to acknowledge responsibility, despite her stated remorse for the errors and her assertion that she had taken steps to address and correct them. Moreover, her other statements to the Court consistently minimize her actions. See Resp. at 9 (“Citation errors, while regrettable, are not uncommon in complex litigation.”); Lindsay Aff. ¶ 6 (“These errors … reflect the inherent challenges of legal research and citation in complex litigation, particularly under time pressure and with large volumes of records.”); Supp. Resp. at 4 (“The acknowledged citation errors were the result of an innocent oversight in the review process, worsened by reliance on error-prone technology.”). Even if she had offered a meaningful apology or accepted responsibility, “regret and apologies are not always enough to avoid sanctions for submitting non-existent legal authority.” The present record shows no genuine acceptance of responsibility….
Lindsay’s bad faith is further evident in the fact that she filed papers containing false citations on two occasions after the Court had issued the OSCs.
First, two months after the initial OSC, she filed a brief in the United States Court of Appeals for the Second Circuit that contained two fabricated citations…
Second, in October 2025, Lindsay submitted objections to a Report and Recommendation entered by the undersigned addressing the motions to dismiss brought by defendants. In a brief opposing the objection, Hirshowitz’s counsel highlighted eight false, fictitious, or “hallucinated” citations — described in detail in the record.
In response, Lindsay stated that she was withdrawing the “specific … citations contained within [the objection] that defense counsel has claimed violate Rule 11” while at the same time asserting that she was doing so “without admission of liability.” In other words, Lindsay again failed to acknowledge her errors or to explain how they occurred. We have reviewed the eight citations in Lindsay’s original objection identified by Hirshowitz’s counsel and agree that, in some instances, they are entirely fabricated.
Her repeated mistakes undermine any claim that the errors were isolated or inadvertent and not an ongoing pattern of disregard for professional obligations. Lindsay asserted that she had taken steps to strengthen her review procedures to ensure greater accuracy going forward. Those unspecified steps have proven to be wholly inadequate….
Deterrence is essential here, given Lindsay’s persistent refusal to explain how the citations were incorporated into her briefs. The Court finds a fine of $2,500 to be an appropriate deterrent, along with an order that she provide her client with a copy of this Opinion and Order. It is also important that other courts learn of Lindsay’s repeated misconduct, which casts doubt on the validity of filings she has made elsewhere. Accordingly, the Court orders Lindsay to deliver a copy of this Opinion and Order to the presiding judge in each pending case where she appears as counsel of record, by filing a letter on the docket and serving opposing counsel….
