Sanction teardown · 3rd Cir. CA, USA · 2025-07-21
McCarthy v. DEA
What happened
In 3rd Cir. CA, 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)Second AI-generated summary of a DEA adjudication relied upon for sanction disparity was inaccurate; counsel admitted and the Court declined to consider it.
-
Misrepresented (Case Law)Third AI-generated summary of a DEA adjudication used to argue inconsistent sanctions was inaccurate; counsel admitted and the Court declined to consider it.
-
Misrepresented (Case Law)Fourth AI-generated summary of a DEA adjudication supporting the 'flagrant departure' claim was inaccurate; counsel admitted and the Court declined to consider it.
-
Misrepresented (Case Law)AI-generated summary of a DEA adjudication cited to show a flagrant departure from past practice was inaccurate; counsel admitted and the Court declined to consider it.
-
Misrepresented (Case Law)Fifth AI-generated summary of a DEA adjudication proffered to show past agency practice was inaccurate; counsel admitted and the Court declined to consider it.
-
Misrepresented (Case Law)Sixth AI-generated summary of a DEA adjudication offered to compare penalties was inaccurate; counsel admitted and the Court declined to consider it.
-
Misrepresented (Case Law)Seventh AI-generated summary of a DEA adjudication invoked for sanction disparity was inaccurate; counsel admitted and the Court declined to consider it.
-
Fabricated (Case Law)AI-generated DEA decision cited as supporting precedent does not exist; counsel admitted and the Court declined to consider it.
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
Relevant pleadings ignored; Order to show cause
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/604/McCarthy_v._US_Mc24-2704-2025-07-21.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).