A Canadian lawyer has been suspended for six months after a judge uncovered ChatGPT-generated fake legal citations in a court hearing.
The Law Society of Ontario’s discipline tribunal suspended Mary Hyun-Sook Lee, who practises as Jisuh Lee, after finding she committed professional misconduct by relying on AI-generated case law that was either fabricated or irrelevant in a family and estate law matter, the National Post reported.
The tribunal ordered Lee to complete remedial measures during her suspension and pay C$10,000 in costs after concluding that she failed to properly represent her client and deliberately misled the court through her submissions.
The disciplinary action follows a hearing in Ontario Superior Court last year, where Justice Fred Myers questioned the validity of several cases cited in Lee’s written submissions. When asked to produce copies of the authorities she relied on, Lee was unable to provide them, prompting the judge to search for the cases himself.
According to the court’s findings, some of the cited decisions could not be found in any legal database, while others were unrelated to the issues before the court or contradicted the arguments Lee advanced. In one instance, a case cited as supporting the removal of an estate trustee was found to have reached the opposite conclusion.
Justice Myers questioned whether artificial intelligence had been used to prepare the material.
In his written reasons, he noted that Lee said her office did not usually rely on such tools but that she would need to confirm with her clerk.
The tribunal found that Lee’s conduct fell below the standard expected of a lawyer, stating that she relied on a factum containing non-existent or irrelevant case law generated by an artificial intelligence tool and misled the court in doing so.
Toronto lawyer Tom Macintosh Zheng, who tracks AI use in legal practice and was cited by the outlet, said the case marks the first time a Canadian law society has suspended a lawyer for misusing artificial intelligence in court submissions, rather than issuing a reprimand.
Zheng added that the case highlights growing concerns within the profession about the risks posed by unverified AI-generated material. “There is increasing scrutiny on how lawyers use these tools, particularly where they affect the integrity of court proceedings,” he said.
The judge emphasised that lawyers have a fundamental duty to verify the authorities they present. He wrote that it should be self-evident that counsel must read and confirm the cases they cite and must not rely on authorities that do not exist or that support the opposite of their position.
Separate civil contempt proceedings against Lee remain ongoing. The court has appointed an amicus curiae, an independent lawyer who assists the court, citing procedural uncertainty, including Lee’s failure to retain legal counsel.
Facts Only
* Mary Hyun-Sook Lee was suspended for six months by the Law Society of Ontario discipline tribunal.
* The suspension followed a finding of professional misconduct regarding reliance on AI-generated case law in a family and estate law matter.
* The tribunal ordered Lee to complete remedial measures and pay C$10,000 in costs.
* A hearing occurred in the Ontario Superior Court last year where Justice Fred Myers questioned cited cases.
* Lee was unable to provide copies of authorities she relied on when asked by the judge.
* Court findings indicated some cited decisions could not be found in legal databases.
* Other cited decisions were unrelated to the court issues or contradicted the arguments advanced.
* One cited case supporting the removal of an estate trustee reached the opposite conclusion.
* Justice Myers questioned whether artificial intelligence was used in preparing the material.
* The tribunal found Lee relied on non-existent or irrelevant case law generated by an AI tool and misled the court.
* Toronto lawyer Tom Macintosh Zheng noted this is the first time a Canadian law society has suspended a lawyer for misusing AI in court submissions.
Executive Summary
A Canadian lawyer, Mary Hyun-Sook Lee, was suspended for six months by the Law Society of Ontario discipline tribunal for professional misconduct. The finding was that she relied on AI-generated case law in a family and estate law matter that was either fabricated or irrelevant. The tribunal ordered Lee to complete remedial measures and pay C$10,000 in costs, concluding that she failed to represent her client properly and deliberately misled the court.
The disciplinary action stems from a hearing in the Ontario Superior Court last year where Justice Fred Myers questioned the validity of cited cases. When asked for the sources, Lee could not provide them, leading the judge to search for them. The court found that some cited decisions could not be located in legal databases, others were unrelated to the case, and one cited decision reached an opposite conclusion to the argument presented. Justice Myers also questioned whether artificial intelligence was used in preparing the materials. The tribunal determined that Lee’s conduct fell below the expected standard for a lawyer by relying on AI-generated, non-existent or irrelevant case law and misleading the court.
This case has drawn commentary from legal observers, noting it as the first time a Canadian law society has suspended a lawyer for misusing AI in court submissions rather than issuing a reprimand. The situation highlights growing professional concerns regarding the risks associated with unverified AI-generated legal material and reinforces the judge's duty for lawyers to verify all cited authorities.
Full Take
The incident reveals a critical tension between the rapid capability of generative AI and the established professional duties of legal practice, specifically the duty to truthfulness and the integrity of judicial proceedings. The pattern observed is one where novel technological tools are introduced into highly regulated professional contexts without sufficient guardrails for verification, leading to systemic risks concerning evidentiary standards. The core implication is a shift in the locus of responsibility: when AI material is introduced, the burden of verifying its authenticity shifts implicitly onto the legal professional.
The narrative suggests that reliance on unverified sources creates an ambiguity where plausible deception (misleading the court) becomes objectively verifiable through subsequent investigation. The reaction from the judiciary and the regulatory body demonstrates a necessary reassertion of foundational legal principles—the duty to verify—against novel shortcuts. The cost imposed and the remedial measures reflect an attempt to recalibrate professional standards when those standards are demonstrably breached by external technological influence.
The pattern points toward a future where legal competence will be measured not just by mastery of law, but by digital literacy and adherence to verifiable sourcing protocols. The emerging challenge is how swiftly legal ethics and procedural rules can adapt to manage the systemic risk posed by tools that can instantaneously generate seemingly authoritative but ultimately hollow information, ensuring that technological advancement does not erode the fundamental integrity of the adjudicative process.
Bridge Questions: How should legal education evolve to integrate AI verification training as a mandatory component? What institutional changes are necessary to establish clear, enforceable standards for the use of AI in legal submissions before further incidents occur? If AI-generated citations become ubiquitous, how will the judiciary develop new methods for assessing the provenance and reliability of submitted legal authority?