Sanction teardown · SC New York, USA · 2026-01-27
Cassata v. Michael Macrina Architect, P.C.
What happened
In SC New York, USA, a filing relied on an unnamed/unconfirmed AI tool to help draft legal argument. The court identified the following problems with the citations in that filing:
-
Misrepresented (Case Law)Counsel cited Becker v Elm Air Conditioning Corp., 143 AD2d 965, 966 (2nd Dep't 1988) for a proposition the case does not support; court found mischaracterization of the holding.
-
Fabricated (Exhibits & Submissions)Counsel copied large portions of an unrelated brief (For a Financial Asset Securitization 2021, LLC v Teona, 2025 WL 334218) that itself had been criticized for non-existent cases; defendant adopted those unverified authorities without verification.
-
Fabricated (Case Law)Counsel cited Harris v Seward Park Housing Corp., 147 AD3d 589 (1st Dep't 2017), which the court found does not exist; citation originated from copied brief and was not verified.
-
Fabricated (Case Law)Counsel cited DiLorenzo v D.C. & D. Transp. Corp., 39 AD2d 950 (2nd Dept, 1972), which the court found does not exist.
-
False Quotes (Case Law)A quotation attributed to Harris v Seward Park Housing Corp. was included in the opposition but does not appear in any cited case and was found only in the copied brief and defendant's filing.
-
Misrepresented (Case Law)Counsel cited New York Univ. v Cont'l Ins. Co., 87 NY2d 308, 323 (1995) for a proposition the case does not support; court found the citation mischaracterized the precedent.
Which AI tool
an unnamed/unconfirmed AI tool. 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
Monetary Sanction; Brief struck (monetary penalty: 10000 USD.)
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/1440/Cassata_v._Macrina_Architects_USA_January_2026.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).