While artificial intelligence (AI) is a tool that can increase lawyers’ efficiency and cut costs for clients, the recent case of Ko v. Li, 2025 ONSC 2965 offers a warning about the serious legal implications for lawyers using artificial intelligence in their practice.
Before using artificial intelligence, lawyers must ensure they adhere to their professional responsibilities and uphold the integrity of the legal profession.
Background
On May 1, 2025, counsel for the applicant, J.L., referenced two non-existent cases while giving her arguments before the Court. The Court discovered the issue after being unable to find any mention of them on any online legal databases. As a result, J.L. was ordered to show cause as to why she should not be in contempt of Court.
In her response, J.L. admitted that the factum had been partially prepared using the generative AI software Chat GPT and that she had neglected to verify the mentioned cases’ accuracy, let alone existence, prior to her submissions.
J.L. explained that the factum in issue had been prepared by members of her staff who had used Chat GPT. However, Justice Myers of the Ontario Superior Court of Justice stated that “the proverbial buck stops with counsel,” meaning that it is up to lawyers to ensure the accuracy of their submissions no matter their level of seniority.
Decision
In his judgment, Justice Myers was clear that “regardless of technology, lawyers cannot rely on non-existent authorities or cases that say the opposite of what is submitted.”
J.L. promptly apologized, withdrew the faulty factum, submitted a corrected version, and committed to undergoing professional development training in legal ethics and technology.
Further, in her May 16, 2025, oral statement, J.L. expressed her regret concerning her past reference to non-existent cases. Acknowledging her responsibility for using these authorities, J.L. emphasized the remedial steps she had taken to rectify her mistake and outlined the new protocols she would implement in her practice to ensure that:
- Every legal authority is independently confirmed through CanLII or Westlaw;
- No AI-generated content will be relied on without rigorous manual review;
- She will no longer delegate legal research to tools she cannot control or verify personally.
Finally, J.L. could not bill her client for the research, factum writing, and attendance from a previous motion.
Conclusion
This case is a prominent wake-up call in the legal field for those using AI in their practices. It drew widespread press, both in general and legal media outlets, and the Court stated that the publicity garnered from the misstep “has served both to publicly denounce inappropriate conduct” and to dissuade others who might lean on AI for legal submissions.
