Sanction teardown · S.D. New York, USA · 2025-11-18
Jorge Paredes Guevara v. A&P Restaurant Corp., et al.
What happened
In S.D. 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:
-
Fabricated (Case Law)Counsel cited Marrero v. Camden Steel Co., 850 F. Supp. 188 (D.N.J. 1994); court identified this authority as non-existent.
-
Misrepresented (Case Law)Counsel cited Baldwin v. Trailer Inns, Inc., 266 F.3d 1104, 1110–11 (9th Cir. 2001) for an FMLA proposition; court noted Baldwin does not mention the FMLA and involves FLSA/breach of contract issues.
-
Fabricated (Case Law)Counsel cited Kennedy v. Cmty. Action Servs., 107 F. Supp. 2d 993 (N.D. Ind. 2000); court identified this authority as non-existent.
-
False Quotes (Case Law)Counsel attributed a non-existent quote to Johnson v. City of Shelby, 574 U.S. 10 (2014); court noted the quoted language does not appear in Johnson.
-
False Quotes (Legal Norm)Counsel quoted non-existent provisions of NYLL § 215 (statutory language the court found was not present in the statute).
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
Not specified in source record.
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/1055/Guevara_v._AP_Restaurant_USA_18_November_2025.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).