Deutsch
English
Español
Français
Nederlands
Português
Canadian Lawyer Suspended for Using ChatGPT to Fabricate Court Cases
A Canadian lawyer has been suspended for six months after using AI-generated fake legal citations in court, marking a landmark disciplinary case over the misuse of artificial intelligence in the legal profession. Toronto attorney Mary Hyun-Sook Lee was found to have submitted nonexistent case law in an inheritance dispute, then compounded the error by blaming a law student who did not exist.
The case, first reported by The Toronto Star, highlights the growing risks of AI hallucinations in legal practice. The growing risks of AI in legal practice have sparked debate among bar associations worldwide.
How the AI Fabrication Was Discovered
In early 2025, during an estate litigation case, Lee submitted written arguments and presented oral arguments citing multiple court decisions that simply did not exist. When opposing counsel challenged the citations, Lee could not produce the actual rulings. Under pressure from the judge, she eventually admitted the documents were generated by ChatGPT, an AI chatbot developed by OpenAI. However, she claimed a student assistant had created them, saying she was "from the old school" and did not understand how ChatGPT worked.
The Law Society of Ontario launched an investigation. Initially, Lee refused to name the assistant, but later confessed she had no associates or interns at the time of the hearing. The fabricated cases included rulings that were entirely invented, and one cited case had a real name but a completely wrong outcome.
Repeated Offense and Digital Trail
Even after being caught, Lee continued to use ChatGPT for legal advice. According to search data retrieved from her computer by the disciplinary tribunal, she asked ChatGPT how to prepare for a follow-up meeting with the judge to avoid a contempt finding and potential disciplinary action. She wrote in the prompt that she had been "so stupid" to rely on nonexistent sources.
The ethical use of AI by lawyers is a growing concern for legal regulators. Lee's case is believed to be the first publicly available disciplinary ruling in Canada specifically addressing AI-generated false citations.
Penalty and Precedent
The Law Society Tribunal imposed a six-month suspension and a fine of 10,000 Canadian dollars (approximately 6,600 euros). The tribunal noted that Lee had shown a lack of candor and had attempted to shift blame. The decision sends a strong message about professional responsibility in the age of AI.
Similar cases have emerged internationally. In the Netherlands, several lawyers received warnings and were ordered to take courses after improperly using AI tools. In the United States, a federal judge sanctioned lawyers in 2023 for citing fake ChatGPT cases in a personal injury lawsuit. The global trend of AI-related legal sanctions is accelerating as courts adapt to new technology.
Implications for the Legal Profession
Legal experts say the case underscores the need for clear guidelines on AI use in law. Many bar associations are now developing rules requiring lawyers to verify all AI-generated content. ChatGPT and similar LLMs can produce highly convincing but entirely fictional legal citations, a phenomenon known as hallucination.
"Lawyers have a fundamental duty of candor to the court. Relying on AI without verification is a serious breach of that duty," said a legal ethics expert quoted in the tribunal documents.
The case also raises questions about the reliability of AI in high-stakes professional environments. As AI tools become more integrated into legal research and drafting, the profession must balance efficiency with rigorous oversight.
Frequently Asked Questions
What happened in the Canadian lawyer AI case?
Toronto lawyer Mary Hyun-Sook Lee was suspended for six months and fined $10,000 CAD for submitting AI-generated fake court cases in an inheritance dispute and then lying about it.
How did the court discover the AI-generated cases?
The opposing counsel challenged the citations, and Lee could not produce the actual rulings. She eventually admitted the documents were created using ChatGPT.
What is an AI hallucination?
An AI hallucination occurs when a large language model generates plausible-sounding but incorrect or nonsensical information, such as fake legal citations.
Are there other similar cases?
Yes. In 2023, U.S. lawyers were sanctioned for citing fake ChatGPT cases. In the Netherlands, multiple lawyers received warnings for improper AI use.
What should lawyers do to avoid AI misuse?
Lawyers should verify all AI-generated content against original sources, disclose AI use when required, and follow bar association guidelines on technology competence.
Follow Discussion