CBA: ISED AI Transparency Submission Urges Nuanced Disclosure

• The Canadian Bar Association (CBA) submitted recommendations to Innovation, Science and Economic Development Canada (ISED) on AI transparency, advocating for a nuanced approach to disclosure.
• The CBA emphasizes distinguishing between AI-generated and AI-assisted content, suggesting disclosure should be mandatory only when AI use is 'sufficiently material' to authenticity.
• Transparency is deemed crucial for content where public reliance on authenticity is high, such as synthetic media depicting identifiable persons or material used in legal contexts.
• The submission reiterates previous CBA recommendations for stronger legal protections against deepfakes, including a federal tort of appropriation of personality or a right of publicity.
• The CBA's consistent stance supports a contextual approach, where AI disclosure obligations are tied to the significance of AI's impact rather than its simple presence.

Lawyers and compliance officers should closely monitor ISED's response to the CBA's recommendations, as they indicate the likely direction of future AI transparency and disclosure regulations in Canada.

The Canadian Bar Association (CBA), through its Privacy and Access Section and Intellectual Property Law Section, has submitted a comprehensive response to Innovation, Science and Economic Development Canada (ISED)'s consultation on AI transparency. Representing over 40,000 legal professionals across the country, the CBA's mandate includes a commitment to improving the law and the administration of justice. The submission highlights the critical need for transparency regarding AI-generated content to foster public trust, while simultaneously cautioning against a blanket approach to disclosure.

The CBA Sections argue that not all uses of artificial intelligence in content creation warrant the same level of scrutiny. Instead, the central question for policymakers should be whether the AI's involvement is sufficiently material to necessitate disclosure. The submission points out that AI tools are increasingly integrated into common applications for drafting, editing, translation, and image enhancement. Mandating disclosure for every instance where AI plays any role could lead to an overwhelming volume of notices, ultimately providing little meaningful information to the public and potentially causing 'notice fatigue,' diminishing the value of disclosures in genuinely important situations.

A crucial distinction underpinning the CBA's recommendations is between content that is entirely AI-generated and content that is merely AI-assisted. This differentiation is key to determining when transparency is truly essential. The need for disclosure becomes particularly strong in scenarios where individuals are likely to rely on content as an authentic record of someone's words, actions, or creations.

Examples provided by the CBA include realistic synthetic audio, images, or video depicting identifiable persons; content that falsely attributes statements or actions to an individual; instances of impersonation or fraud; false content with the potential for substantial personal impact; deceptive commercial representations; and any material used in legal, governmental, electoral, or other contexts where authenticity is a core expectation. Conversely, minor AI assistance, such as basic editing, formatting, or accessibility enhancements that do not materially alter the substance or authenticity of the content, should generally not trigger the same disclosure obligations as synthetic content designed to mislead the public.

Deepfakes, which represent a significant concern for content authenticity, receive particular attention in the CBA's submission. Building on its October 2025 submission to Canada's AI Strategy, known as the National Sprint Submission, the CBA reiterates its recommendation for the federal government to consider stronger legal protections against the unauthorized commercial exploitation of a person's likeness or voice through AI-generated deepfakes. This includes proposing the establishment of a federal tort of appropriation of personality or, for commercial artists, a federal right of publicity.

Furthermore, the CBA suggests that Canada explore protections for personal attributes such as an individual's face and voice. These recommendations underscore the growing legal implications of advanced AI technologies and the potential for misuse, particularly in creating deceptive content that could harm individuals or undermine public trust. The focus on deepfakes highlights the urgent need for legal frameworks that can adapt to rapidly evolving AI capabilities.

The contextual approach to AI disclosure advocated by the CBA is consistent with its previous work. In its May 2025 submission concerning immigration modernization, the association recommended a mandatory disclosure rule specifically for instances where AI is utilized to create pleadings or evidence presented before the Immigration and Refugee Board. This recommendation was carefully distinguished from the use of non-essential administrative technologies, further illustrating the principle that disclosure should be tied to the significance of AI's impact rather than its mere presence.

Lawyers and compliance officers should closely monitor ISED's response to the CBA's recommendations, as they indicate the likely direction of future AI transparency and disclosure regulations in Canada. These potential regulations will significantly impact compliance obligations for AI tool usage, particularly concerning content authenticity in legal, commercial, and public reliance contexts. The discussions also suggest the possibility of new torts being developed to address the misuse of likeness and other personal attributes through AI technologies.

Lawyers and compliance officers should closely monitor ISED's response to the CBA's recommendations, as they indicate the likely direction of future AI transparency and disclosure regulations in Canada. This will impact compliance obligations for AI tool usage, particularly regarding content authenticity in legal, commercial, and public reliance contexts, and could lead to new torts for misuse of likeness.

Source: Reporting based on the Canadian Bar Association's AI transparency submission.