What was hallucinated

Fabricated: Case Law | Fabricated precedent case cited in Ms. Lee’s factum (one of several); the judge could find no record of it and Ms. Lee later confirmed it was an AI-generated hallucination from ChatGPT; the factum was revised to remove fake citations. || Fabricated: Case Law | Non-existent precedent case relied upon orally at the May 1, 2025 hearing (second of two); the court could not locate it and Ms. Lee admitted it was an AI hallucination. || Fabricated: Case Law | Non-existent precedent case relied upon orally at the May 1, 2025 hearing (first of two); the court could not locate it and Ms. Lee admitted it was an AI hallucination.

Details

(Order to show cause is here.)At the end of the show cause proceedings, Justice Myers noted that, due to the media reports about this case, the goals of any further contempt proceedings were already met, including: "maintaining the dignity of the court and the fairness of civil justice system, promoting honourable behaviour by counsel before the court, denouncing serious misconduct, deterring similar future misconduct by the legal profession, the public generally, and by Ms. Lee specifically, and rehabilitation".The judge therefore declined to impose a fine or to continue the contempt proceedings, on the condition that Counsel undertakes Continuing Professional Development courses (as she said she would), and does not bill her client for any unrelated work (which was helped by the fact that she had so far been working pro bono).SequelIt later surfaced that Ms. Lee had not been fully honest with the court, leading to renewed contempt proceedings (see here). In July 2026, the Law Society of Ontario suspensed her for six months, adding a 10,000 CAD fine (see here).

Sanction teardown · Ontario SCJ, Canada · 2025-05-28

Ko v. Li

What happened

In Ontario SCJ, Canada, a filing relied on ChatGPT to help draft legal argument. The court identified the following problems with the citations in that filing:

Fabricated (Case Law)

Fabricated precedent case cited in Ms. Lee’s factum (one of several); the judge could find no record of it and Ms. Lee later confirmed it was an AI-generated hallucination from ChatGPT; the factum was revised to remove fake citations.
Fabricated (Case Law)

Non-existent precedent case relied upon orally at the May 1, 2025 hearing (second of two); the court could not locate it and Ms. Lee admitted it was an AI hallucination.
Fabricated (Case Law)

Non-existent precedent case relied upon orally at the May 1, 2025 hearing (first of two); the court could not locate it and Ms. Lee admitted it was an AI hallucination.

Which AI tool

ChatGPT. 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

Plaintiff’s application dismissed; no costs imposed; court warns against future use of generative AI without verification; subsequent suspension and fine (monetary penalty: 10000 CAD.)

Additional detail

(Order to show cause is here.)At the end of the show cause proceedings, Justice Myers noted that, due to the media reports about this case, the goals of any further contempt proceedings were already met, including: "maintaining the dignity of the court and the fairness of civil justice system, promoting honourable behaviour by counsel before the court, denouncing serious misconduct, deterring similar future misconduct by the legal profession, the public generally, and by Ms. Lee specifically, and rehabilitation".The judge therefore declined to impose a fine or to continue the contempt proceedings, on the condition that Counsel undertakes Continuing Professional Development courses (as she said she would), and does not bill her client for any unrelated work (which was helped by the fact that she had so far been working pro bono).SequelIt later surfaced that Ms. Lee had not been fully honest with the court, leading to renewed contempt proceedings (see here). In July 2026, the Law Society of Ontario suspensed her for six months, adding a 10,000 CAD fine (see here).

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/451/Ko_v._Li_Ontario_SC_Canada_May_28_2025.pdf, via Damien Charlotin's public AI Hallucination Cases Database (CC0).