Canadian Judicial Council: Releases AI Digital Evidence Guidelines

• The Canadian Judicial Council has released new guidelines to help federal judges navigate AI tools and evaluate digital media evidence, including deepfakes.
• The 2026 AI guidelines encourage judges to leverage AI for justice administration while adhering to judicial principles, outlining appropriate and inappropriate uses.
• A separate set of guidelines provides checklists for judges to assess acknowledged, unacknowledged, and indirect AI evidence.
• Individual judges using AI for administrative tasks typically do not need to disclose its use, similar to other support tools.
• Courts, as institutions, must be transparent about their AI systems, including their evaluation, governance, security, and oversight mechanisms.

Courts, as institutions, are mandated to be transparent about their AI tool usage, including how these systems are evaluated, approved, governed, and secured, alongside the safeguards and oversight mechanisms in place.

The Canadian Judicial Council (CJC), the body responsible for overseeing Canada's federal judiciary, has released two significant sets of guidelines aimed at assisting judges in navigating the complexities of artificial intelligence (AI) and evaluating digital media evidence. These new directives are designed to help the Canadian federal judiciary effectively manage AI tools and address the challenges posed by digital evidence, including the detection of deepfakes.

One set of guidelines, specifically the 2026 AI guidelines, focuses on encouraging judges to strategically employ AI to support the administration of justice. This encouragement is coupled with a strong emphasis on ensuring that any AI use aligns with established judicial principles. The second set of guidelines addresses the critical task of evaluating digital media evidence, offering judges practical assistance in cases where evidence is suspected or alleged to have been created or altered by AI tools. This includes providing checklists to aid in the assessment of acknowledged, unacknowledged, and indirect AI evidence.

These updated documents build upon the CJC's initial AI guidelines, first issued in 2024. The earlier 2024 guidance primarily aimed to raise awareness regarding the risks associated with AI while simultaneously encouraging the judiciary to explore safe, effective, and appropriate applications of AI tools. Multiple courts across Canada have indicated that these foundational guidelines have already influenced their judges' approaches to generative AI technologies.

The shift in focus from the 2024 guidelines to the 2026 iteration signifies an evolution in the CJC's stance on AI. Tom Macintosh Zheng, a former commercial litigator whose organization, Courtready.ca, monitors court cases involving AI-generated case citations, highlights this change as particularly noteworthy. He observes that the 2026 AI guidelines provide clear direction on appropriate and inappropriate uses of AI for judges, with an appendix detailing specific examples of suitable, high-risk, and unsuitable applications.

Macintosh Zheng views this progression from risk awareness to actively encouraging AI adoption, albeit with explicit guardrails, as a crucial development. He suggests that AI tools possess the potential to significantly enhance the efficiency of court operations, questioning why the benefits of artificial intelligence should be confined solely to the private sector. He anticipates that judges and court administrators, upon reviewing these guidelines, will become more adept at identifying areas where AI can genuinely offer advantages to judicial practice.

Kathryn Manning, who was a founding partner at DMG Advocates, underscores the timeliness of these guidelines, particularly those pertaining to digital media evidence. She points out that the rapid advancement of AI technologies has introduced new complexities in verifying the authenticity of evidence. Manning emphasizes that lawyers seek assurance that judges possess the necessary tools to navigate this swiftly changing technological landscape, acknowledging that expertise in rapidly evolving technology was not a traditional part of a judge's daily responsibilities just a few years ago.

A key aspect of the new AI guidelines concerns disclosure requirements for AI tool usage. Generally, an individual judge's use of AI tools for administrative, organizational, or editorial assistance will not typically necessitate disclosure. This is consistent with the non-disclosure of other common support tools and personnel, such as law clerks, legal researchers, translation software, or spelling and grammar checkers. In such instances, the judge retains sole responsibility for the work produced.

However, a distinct and more stringent requirement applies to courts as institutions. Courts, as institutions, are mandated to be transparent about their AI tool usage, including how these systems are evaluated, approved, governed, and secured, alongside the safeguards and oversight mechanisms in place. This institutional transparency is crucial for explaining the types of functions AI systems perform and for ensuring that their use remains consistent with fundamental principles such as judicial independence, procedural fairness, and the proper administration of justice.

Lawyers and compliance officers must familiarize themselves with these new guidelines to understand how Canadian federal judges will evaluate AI-generated or deepfake evidence, and to anticipate judicial expectations and disclosure requirements regarding the use of AI tools in legal proceedings and court administration.

Source: Original reporting via Canadian Lawyer Magazine