Sanction teardown · TJ Rio de Janeiro, Brazil · 2026-05-06
Thalita Aparecida Araújo Rosa Campos v. Marcio Abreu Rosa
What happened
In TJ Rio de Janeiro, Brazil, 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)Appellee's filings included citations to precedents attributed to state courts that the tribunal found to be nonexistent; counsel later admitted the references resulted from AI-assisted research but failed to verify them.
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
Monetary Fine; Bar Referral (monetary penalty: 1 BRL.)
Additional detail
The appellate court determined that the appellee's counsel cited non-existent state-court precedents in pleadings. The counsel admitted the citations arose from inexperience using AI research tools but the court held she had a duty to verify results and characterized the conduct as litigância de má-fé. The court imposed a 2% procedural fine on the updated value of the claim, ordered referral to the OAB-RJ for possible professional measures, and rejected the excuse of mere AI error as excusing the behavior.
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/2878/www3.tjrj.jus.br_EJURIS_temp_f9976351-a587-4d57-92b0-85cb9fa10198.html.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).