Sanction teardown · CA Oregon, USA · 2026-04-22
Carol L. Williams v. Tracy Honl
What happened
In CA Oregon, 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)Brief cited 'Tubra v. Cooke, 233 Or App 339, 225 P3d 862 (2010)' for an anti‑SLAPP proposition; court found the cited case does not exist.
-
Misrepresented (Case Law)Brief attributed de novo‑review language and standards to Neumann v. Liles; court found Neumann does not state the claimed proposition.
-
Misrepresented (Case Law)Brief quoted Staten v. Steel as supporting a broad rule that private employment disputes are not matters of public interest; court found Staten does not say that.
-
False Quotes (Case Law)Brief relied on Davoodian v. Rivera and similar citations for propositions or quotations the court identified as omitted or unsupported; listed among problematic citations.
-
Misrepresented (Case Law)Brief invoked Handy v. Lane for a proposition about government transparency and accountability that the court found Handy did not contain or emphasize as claimed.
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
Brief Struck; Adverse Costs Order (monetary penalty: 8044 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/2017/Williams_v._Honl_USA_22_April_2026.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).