Sanction teardown · Haifa Magistrate's Court, Israel · 2025-06-08
Chen v. Vana et al.
What happened
In Haifa Magistrate's Court, Israel, 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)Authorities cited in section 28 of the appeal were nonexistent; the trustee noted they do not appear in Nevo, the Court independently failed to locate them, ordered production, and the appellant admitted they were AI-generated. The Court noted SC guidance that such use can justify dismissal and imposed costs.
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
Appeal was ultimately dismissed on merits; Monetary sanction (monetary penalty: 3000 ILS.)
Additional detail
The appellant, a lawyer representing himself in an appeal, submitted a brief that included non-existent legal citations. An employee in the lawyer's office had used an AI system to assist in drafting the document. After the opposing counsel and the court itself were unable to locate the cited precedents, the lawyer admitted they were AI-generated and a "good-faith mistake" for which he took full responsibility. While citing its authority to dismiss the appeal entirely due to the submission of non-existent sources, the court chose a "softer" sanction. It proceeded to hear the case, ultimately dismissing the appeal on its merits and imposing a separate monetary penalty payable to the State Treasury for the misconduct.
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/496/Chen_v._Vana_et_al._Israel_8_June_2025.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).