Jisuh Lee referred for criminal contempt over AI-generated fake citations in Ontario court
Ontario lawyer Jisuh Lee submitted a factum with hallucinated or misattributed citations generated by ChatGPT. After initially denying AI involvement, she admitted using AI, and a court referral to the Attorney General followed for potential contempt.
You can delegate drafting to an algorithm, but you cannot delegate your integrity.
Key facts
- What
- Ontario lawyer Jisuh Lee submitted a factum with hallucinated or misattributed citations generated by ChatGPT.
- Incident date
- May 1, 2025
- Who
- Jisuh Lee
- Failure mode
- Hallucination
- AI surface
- Chatbot
- Severity
- Catastrophic
What happened
In Ko v. Li, 2025 ONSC 6785, attorney Jisuh Lee used ChatGPT to draft a legal factum that included four non-existent case citations. When confronted by Justice Fred Myers, Lee denied AI involvement and claimed a student did the work. She later admitted to the court that there was no student and that she had used ChatGPT to draft the factum.
What broke inside the model
- 01 · TriggerA user asks for a fact, a citation, or a figure.
- 02 · Model stepThe model writes a fluent, confident answer.
- 03 · Control gapNothing ties the claim back to a real source.
- 04 · FailureA fabricated fact ships as if it were verified.
- 05 · ConsequenceThe false claim reaches a customer, a court, or the public.
Confidence holds, and even spikes, as the claim detaches from any source.
The LLM hallucinated legal precedents, creating citations that appeared authentic but were non-existent. The failure was compounded by the lawyer's failure to verify the AI output and subsequent attempts to deceive the court.
What it cost
Sources
Cite this entry
https://failureindex.ai/failures/jisuh-lee-referred-criminal-contempt-generatedAI Failure Index. "Jisuh Lee referred for criminal contempt over AI-generated fake citations in Ontario court" (FI-0270). Realm Labs. https://failureindex.ai/failures/jisuh-lee-referred-criminal-contempt-generated (indexed Jun 5, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0270. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
- AI Detection & Response (AIDR)
A runtime layer that watches the model's internal state can flag the moment a model commits to a claim it has no support for, and hold or reroute the response before it reaches a user. Realm reads those signals in real time rather than grading the transcript after the fact.