Artificial intelligence (“AI”) is changing the legal landscape. AI can assist with research, drafting and even formulating legal arguments. However, as recently cautioned by Justice Griffin in the BC Court of Appeal, when used improperly, AI can undermine a litigant’s position and even increase their costs.
1182573 B.C. Ltd. v Gerber, 2026 BCCA 399 involves disputes arising from a purchase and sale agreement between the seller/appellant, Halfmoon Waterfront Properties Inc. (“HWPI”), and the buyers/respondents, Martin and Michelle Gerber (the “Respondents”). HWPI assigned its rights and causes of action arising out of the agreement to 1182573 B.C. Ltd (collectively, the “Appellants”).
In March 2026, the Respondents commenced an application seeking certain relief, such as an order for security for costs and to substitute one Notice to Admit for the two duplicative ones. In April 2026, Justice Elwood ordered the Appellants to pay $20,000 as security for costs and provided directions regarding a revised Notice to Admit (the “Elwood Order”). The Appellants sought to appeal the Elwood Order without seeking leave. The Respondents applied to quash the appeal on the basis that the Appellants required leave, which the Appellants argued was not needed.
The Court held that leave was required and exercised its discretion to convert the notice of appeal to a leave application, notwithstanding the Appellants’ failure to seek leave. The Court ultimately declined to grant leave, finding that the appeal was disproportionate to the issues in dispute and that allowing the appeal would run contrary to the interests of justice.
This blog will not delve into the analysis of the Court’s decision on declining leave but will rather focus on Justice Griffin’s advice to self-represented litigants on the use of AI. The Appellants were not represented by counsel, but rather by their sole shareholder and director, Mr. Alister Toma. Mr. Toma is a real estate developer and not a lawyer. He admitted to the Court that he relied on AI in preparing his submissions. Such an admission was not a surprise to the Court, as Mr. Toma’s pleadings, correspondence and submissions appeared to be AI-generated.
Justice Griffin cautioned that without sufficient human oversight, generative AI tools can undermine a litigant’s position. Firstly, AI may ‘hallucinate’ by making up fictitious case law or by incorrectly referencing a real case. A litigant who files submissions with false authority violates the BC Registrar’s Filing Directive, which explicitly states that litigants are responsible for the authenticity and accuracy of all materials filed with the Court. Similar safeguards are in place through Ontario’s Civil Practice Direction, which requires litigants to guarantee the accuracy of their materials. Litigants who file inaccurate materials risk a costs award being issued against them while also jeopardizing their credibility with the Court.
Another issue for self-represented litigants using AI, especially those without legal training, is the risk of referencing irrelevant authorities. When generating submissions, AI tends to generate long lists of authorities and extract passages from these authorities to support a position. Without proper legal training, self-represented litigants may not realize that many of the cited authorities are irrelevant to the issues before the Court. Such submissions are incoherent and undermine a litigant’s case, as it is the “opposite of persuasive advocacy” to cite irrelevant authorities.
AI may also give a litigant bad legal advice. In Justice Griffin’s observations, AI “can be quite delusional about one’s legal rights and how to present legal argument.” AI-generated material that is unreviewed by someone with legal training tends to be unduly lengthy and verbose and to contain ‘legal’ language untethered from established principles. Furthermore, social and legal systems are complex. Understanding the law and exercising good judgement to advance a position is an art that cannot simply be recreated. AI cannot fully grasp these nuances and cannot replace the years of experience of professionally trained lawyers. Unlike self-represented litigants, lawyers who use AI have the foundational knowledge and judgement to use the tool effectively and offset any risks.
AI-generated materials also lack human morals. Lawyers are taught to be honest, ethical, and polite. Such moral standards are absent from AI, which often leads to poor conduct. This not only results in heightened conflict and costs for litigants but also undermines their credibility with the Court.
Finally, while many self-represented parties use AI as a cost-saving measure, the use of AI without independent and informed judgement may actually increase costs. As stated above, a judge may order costs against a litigant relying on false AI-generated content. AI has also increased the number of active legal proceedings, many of which may be baseless or without merit. The increase in the number of legal proceedings puts a strain on our limited and valuable judicial resources. Judges “are a limited human resource” that must organize their time to best serve the broader public. The rise of meritless litigation fuelled by AI diverts a judge’s attention away from other matters and wastes valuable judicial resources.
As stated by Justice Griffin, “a litigant who decides to use AI to help them in their case needs to be cautious, and to be aware that it might not help but rather hinder and lead them astray. They could end up disappointed and confused, as well as subject to successive costs awards.” With the great powers of AI comes great responsibility. Absent foundational legal knowledge and experience, litigants must be aware of the risks of relying on AI to act as their ‘lawyer.’
SUBSCRIBE TO OUR BLOG FOR THE LATEST FROM DE VRIES
