Sanction teardown · TJ San Catarina, Brazil · 2025-09-16
Agravo de Instrumento nº 5059805-76.2025.8.24.0000
What happened
In TJ San Catarina, 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)Cited several non-existent STJ and state-court precedents (process numbers that do not correspond to real decisions); court found these authorities fabricated.
-
False Quotes (Case Law)Attributed a passage to AgInt no AREsp 1.518.381/RJ that the court found absent from that decision (false quotation / misattribution).
-
False Quotes (Doctrinal Work)Claimed a specific sentence appears on page 284 of Humberto Theodoro Júnior's Curso de Direito Processual Civil (66ª ed.); court could not find that quotation in the cited doctrinal source.
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 (monetary penalty: 1 BRL.)
Additional detail
The relator identified that the appellant-lawyer cited multiple jurisprudential authorities and a doctrinal passage that do not exist or do not contain the quoted text. The court concluded these were fabricated or misattributed authorities — likely produced by indiscriminate use of AI — and held the behavior amounted to litigância de má-fé (breach of duties in arts. 77 and 80 CPC). As sanction for the hallucinated/fictitious citations the court imposed a monetary penalty for litigância de má-fé and rejected the agravo interno. The opinion stresses the duty of candor and caution when using AI and the need to verify authorities.
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/2860/Jurisprud%C3%AAncia_Catarinense_-_TJSC.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).