Sanction teardown · GAO, USA · 2025-09-25
Oready, LLC (2)
What happened
In GAO, 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)Protester cited Trailboss Enterprises, B-417212.2; GAO found no such decision exists and protester’s 'typo' explanation was unpersuasive.
-
Fabricated (Case Law)Protester cited 'JRS Staffing Services, B-409407.7, Apr. 2, 2014, 2014 CPD ¶ 116'; GAO found no such B-number and the CPD ¶116 actually corresponds to a different decision (Government and Military Certification Sys., B-409420).
-
Misrepresented (Case Law)Protester cited Kathpal Technologies, Inc., B-291637.2 claiming it sustained a protest about letters of intent; GAO found the Kathpal decision exists but does not address letters of intent and in fact denied a post-award past-performance challenge.
-
Fabricated (Case Law)Protester cited System Studies & Simulation, Inc., B-416693; GAO found no such decision exists.
-
Fabricated (Case Law)Protester cited JRS Staffing Services, B-409407 to support claim about unstated evaluation criteria; GAO found no decision using that party with B-409407 (B-409407 corresponds to JCMS) and the cited proposition is unsupported.
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
Protests dismissed for abuse
Additional detail
Actually the fourth order in that case that pertains to hallucinations; a first, June 5 Order is only mentioned in a second, June 18 order that does not call out what appears to be hallucinated references.
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/815/Oready_881816.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).