Sanction teardown · E.D. Michigan, USA · 2025-09-25
Evans, et al. v. Robertson et al. (2)
What happened
In E.D. Michigan, 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:
-
False Quotes (Case Law)Evans misattributed quotations to United States v. Ziegler Bolt & Parts Co., 111 F.3d 878 (Fed. Cir. 1997); court flagged the inaccuracy.
-
False Quotes (Case Law)Evans misattributed quotations to Sec. & Exchange Comm'n v. Knowles, 87 F.3d 413 (10th Cir. 1996); court flagged the inaccuracy.
-
Fabricated (Case Law)Robertson cited Mickens v. 10th Jud. Dist. Ct., 460 F. App'x 728 (9th Cir. 2011); court found the citation inaccurate and noted the actual related decisions were Eleventh Circuit cases addressing different issues.
-
Fabricated (Case Law)Evans cited Aguirre v. Reese, No. 2:12-cv-01837, 2013 WL 6577085, at *5 (D. Nev. Dec. 13, 2013); court found the citation nonexistent and the case number assigned to a different matter.
-
False Quotes (Case Law)Evans attributed the quotation “Parties may not evade service by hiding behind pseudonyms while continuing to engage in legal actions using those same identities.” to Doe v. F.B.I., 218 F.R.D. 256 (D. Colo. 2003); court found no such language in that case or elsewhere.
-
Fabricated (Case Law)Robertson cited McNair v. Everly, No. 18-12511, 2019 WL 2560069 (E.D. Mich. June 20, 2019); court determined that citation does not exist and the case number corresponded to a different matter.
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
Warning to both parties
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://reason.com/volokh/2025/09/25/ive-seen-fake-cites-on-both-sides-now/, via Damien Charlotin's public AI Hallucination Cases Database (CC0).