\n\n
Key developments: A British Columbia Court of Appeal justice warned that AI‑prepared submissions are overwhelming judges, who "cannot provide an unlimited audience" to irrelevant, verbose filings. The ruling highlights systemic risks such as hallucinated cases, irrelevant citations, and overly aggressive tone in AI‑generated pleadings. Similar concerns echo across Canada, where surveys show over half of self‑represented litigants now use AI tools, and a sizable share of family‑law and Supreme Court leave applications involve self‑representation.
Why it matters: Judicial resources are finite, and the surge in AI‑driven, lengthily‑crafted documents forces courts to devote time to filtering out nonsense rather than adjudicating substantive disputes. Experts like Amy Salyzyn note a "significant new burden" on workloads, while Justice Griffin cautions that reliance on AI can "hinder and lead them astray," potentially leading to cost sanctions for litigants.
What to watch next: • Further court rulings on AI‑generated filings, especially any standards for disclosure or sanctions. • Ongoing copyright battles such as the US Third Circuit decision that AI training on copyrighted Westlaw headnotes is not fair use, signaling tighter limits on AI data use. • International guidance, notably UNESCO’s 2025 AI‑in‑courts guidelines, which may shape domestic rule‑making on transparency, auditability and human oversight. • Monitoring of self‑represented litigants’ AI usage rates and any legislative or policy responses aimed at balancing access to justice with preserving judicial efficiency. [[1]]