Compiled by Citation Safe (Digital Empire Holdings LLC) from the court’s published sanction order (decided September 14, 2026). Quotes and case details come from the order itself — check the primary source before relying on them.

Sanction teardown · S.D.N.Y., USA · 2026-09-14

Tiekert v. Village of Mamaroneck, et al.

What happened

In S.D.N.Y., USA, a filing relied on ChatGPT to help draft legal argument. The court identified the following problems with the citations in that filing:

False Quotes (Case Law)

Counsel attributed a quotation to Back, but the Court could not locate it; counsel also claimed Back held that officials may be liable for carrying out discriminatory or retaliatory actions initiated by others. The Court found that Back instead held the superintendent lacked personal involvement and undermined the cited proposition.
Misrepresented (Case Law)

Counsel stated that Mangino sustained a First Amendment retaliation claim based on selective enforcement. The Court explained that Mangino affirmed dismissal of the claim and did not sustain it.
Misrepresented (Case Law)

Counsel attributed to Dolan a discussion of inferring retaliatory animus from temporal proximity and related circumstantial evidence. The Court found that Dolan did not discuss that principle.
False Quotes (Case Law)

Counsel quoted Gorman-Bakos as stating that close temporal proximity suffices to plead causation. The Court could not locate that quotation, although it fairly summarized principles in the opinion.
False Quotes (Case Law)

Counsel attributed to Dorsett a quotation concerning temporal proximity and retaliatory animus defeating dismissal. The Court could not locate the quotation and noted that Dorsett addressed the injury element rather than causation.
False Quotes (Case Law)

Counsel attributed to Dillon a quotation stating that causation is sufficiently pleaded where the complaint describes a chronology suggesting retaliation. The Court could not locate the quotation.
False Quotes (Case Law)

Counsel attributed to Ragbir a quotation concerning pleading retaliatory animus without direct proof. The Court could not locate the quotation and noted that Ragbir’s judgment had been vacated in light of later Supreme Court authority, a subsequent history counsel failed to disclose.
Misrepresented (Case Law)

Counsel cited Dougherty and Davis for qualified-immunity propositions, but neither case mentioned qualified immunity. Counsel also included a partial quotation attributed to Dougherty that the Court could not locate.
Misrepresented (Case Law)

Counsel cited Mangino for the proposition that qualified immunity could not be resolved without factual development. The Court explained that Mangino addressed qualified immunity in an abuse-of-process claim and upheld dismissal on qualified-immunity grounds.
Misrepresented (Case Law)

Counsel cited Kerman for the proposition that clearly established law prohibited retaliation for protected speech. The Court noted that Kerman analyzed qualified immunity in unlawful-seizure and false-imprisonment claims, not retaliation for protected speech.
Misrepresented (Case Law)

Counsel cited Gonzalez for the proposition that qualified immunity could not be resolved on a motion to dismiss where pleaded facts contradicted the defense. The Court explained that Gonzalez involved summary judgment and affirmed dismissal on qualified-immunity grounds.

Which AI tool

ChatGPT. Note: Charlotin's public database records tool attribution only where a court order, brief, or reporting on the matter states it explicitly; "unidentified" or "implied" means the record indicates AI use but does not name a specific product — we do not guess.

Outcome

Attorney Brian M. Higbie was found to have violated Rule 11 and sanctioned $500, required to complete two New York CLE credits on AI ethics, disclose future GenAI use in filings, and serve the order on his client. (monetary penalty: 1 .)

Additional detail

Plaintiff’s counsel admitted relying on ChatGPT to research and draft opposition briefs, verifying only that cited cases existed, were from the correct jurisdiction, and concerned the general subject matter. The Court found that the briefs contained nonexistent quotations and materially misrepresented the holdings of multiple real cases, including Back, Mangino, Dorsett, Dillon, Ragbir, Mangino, and Gonzalez. The Court held that counsel’s failure to read and verify the authorities violated Rule 11 and constituted bad faith, emphasizing that prior awareness of ChatGPT’s fabrication risks and counsel’s failure to investigate after being alerted supported conscious avoidance. The Court imposed a $500 sanction payable to the Clerk, along with nonmonetary safeguards, but denied defense requests for attorney’s fees because no separate Rule 11 motion had been filed.

How Citation Safe would have caught this

Citation Safe runs three deterministic layers before a brief is filed: (1) does the citation exist against CourtListener's database of published opinions, (2) if quoted, does that exact language appear in the source, (3) does the cited case actually support the proposition it is cited for. Fabricated case citations fail Layer 1. Fabricated or misattributed quotations fail Layer 2 even when the underlying case is real. Misrepresented holdings — a real case cited for a proposition it does not support — are the target of Layer 3. None of these checks involve asking another language model whether the citation looks right; they are lookups and text-matches against the actual source, which is why a hallucinated citation has to survive a direct lookup against the authoritative source — not another model's opinion — to earn a VERIFIED stamp; our measured false-verify rate is published live at /quality.

Check a brief before you file it → · See our live false-verify rate

Source: https://www.damiencharlotin.com/documents/3068/Tiekert_v._City_of_Mamaronek_USA_14_September_2026.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).