The use of artificial intelligence (“AI”) is exploding in Canada and around the world, making it a key focus across political, legal, and business fields. At the same time, more employees are using AI in their work, often without guidance or supervision. Like all emerging technologies, AI is a powerful tool with both significant opportunities and risks, which can be mitigated through carefully crafted workplace policies.
Recent AI incidents
Canadian courts have turned their attention to the use of AI, reflected in several recent decisions regarding the misuse of AI in court submissions. The most recent of these is the Federal Court decision, Lloyd’s Register Canada Ltd. v. Choi, 2025 FC 1233 (“Lloyd’s”). This serves as a cautionary tale for would-be litigants, highlighting the serious nature of misleading courts with AI-generated materials.
The case involved an application by Lloyd’s to remove a motion filed by defendant Munchang Choi that contained reference to a non-existent incident, namely “AI hallucinations”. Mr. Choi, who identified himself as such, claimed that the use of AI tools was limited to drafting and research, and that he made a mistake when transcribing references to the incident. However, this is not the first time Choi has relied on AI-generated cases, and applicants have raised concerns about the reliability of his explanations.
The court found that the use of the “hallucination” case was a serious issue, and that Mr. Choi was not fully responsible for his actions to mislead the court and had not expressed adequate remorse. The court ordered that the record of the complaint be removed from the court’s file and awarded the applicant costs.
In Ko v Lee, 2025 ONSC 2965, the applicant’s attorney, Jisuh Lee, referred to several non-existent or “spurious” court precedents generated through ChatGPT in his argument.
The court found that Mr Lee had breached his duties to the court and stressed that the lawyer’s misrepresentation of the law posed a real risk of miscarriages of justice that undermined the dignity of the court and the fairness of the civil justice system. Ms. Lee’s subsequent reflective response in court has mitigated the potential consequences that could be imposed in this case.
The significant publicity surrounding this case condemned the misuse of AI in Canadian courts and acted as a deterrent to the legal profession, reminding lawyers of the serious consequences that can result from trusting AI-generated submissions without first verifying them.
Finally, Zhang v Chen, 2024 BCSC 285 also involved the use of ChatGPT by lawyers in the preparation of court documents and the careless use of “fake” lawsuits. Mr. Ke, the defendant’s lawyer, mistakenly included two fictitious AI-generated cases in his client’s application without verifying them. Mr. Zhang demanded that Mr. Ke pay for the time he spent responding to a case that did not exist.
The court found that Mr. Ke did not include the proceedings with deceptive intent and that the circumstances of the proceedings did not justify an award of special costs. She had expressed regret that she had used a generative AI tool that was not fit for legal purposes and had been exposed to significant negative publicity. However, the court found that Ms. K was personally liable for the costs associated with the additional effort and expense caused by the disruption caused by the AI’s hallucinations.
Apart from these major decisions, the issue of AI hallucination cases has been raised in various legal contexts, including courts and criminal proceedings. These cases often involve self-styled litigants with limited resources and legal expertise who may not be aware of the potential for AI to generate non-existent cases.
Mitigating AI risks through workplace policies
The examples above highlight the importance of companies proactively managing and avoiding legal and reputational risks by establishing policies and procedures for the responsible use of AI by employees. These policies should address issues such as the types of AI platforms that can be used, privacy and information security, responsibility for overseeing the use of AI and verifying the source, intellectual property (IP) infringement, and other related issues that may arise in the specific context of your business.
Businesses must also ensure compliance with new laws targeting AI. For example, in Ontario, the Labor Force Act 149 of 2024 introduces new requirements to disclose the use of AI in the recruitment process for public and online jobs, effective January 1, 2026. For more information on this requirement, see Mira Nemr’s recent article, “Employer Considerations for Job Postings: The Impact of Bills 149 and 190.”
Our employment law team can help
If you have any questions about your AI workplace policy or need help drafting one, please contact our employment law team.
About the author
Kate Tello is Research Director at Parley Robertson Hill & McDougall LLP/srl. He was called to the Ontario Bar in 2006 and brings extensive experience providing legal, policy and governance advice across industries including government, health care, finance and labor. Kate is known for her strategic research insights, leadership skills, and ability to help clients effectively manage legal and reputational risks.
