Controls on generative AI in Canada
National laws specifically addressing AI have not yet passed in Canada. However, Bill C-34 (if enacted) would impose specific obligations on operators of ‘regulated chatbot services’, including requirements to mitigate the risk of communicating harmful content and to address specifically identified harmful behaviours. These provisions are discussed in the User transparency section. Canada’s export control regime is primarily based on the multilateral Wassenaar Arrangement, which does not itself explicitly list AI in current control lists (though high-performance computing systems, encryption tools or network intrusion software, or certain imaging or machine vision sensors that may form part of AI technologies may meet criteria for control).
Due to stalls in global consensus on updating the Wassenaar Arrangement, on 20 July 2024, Canada unilaterally added certain quantum computing and advanced semiconductor technologies to its Export Control List, effectively prohibiting their export to any location other than the United States without an export permit. The list of controlled goods specifically added is part of Export Control List Order SOR/2024-112, where the attached Regulatory Impact Analysis Statement mentions specifically the addition of gate-all-around field-effect-transistors/GAAFET based on their application in creating microchips that run faster and consume less power, thus enabling more powerful and efficient artificial intelligence applications, including for military systems.
Under Canada’s national security powers under the Investment Canada Act, it is advisable to work with counsel to develop a strategy for managing the requisite notification to and/or review by government for all new businesses or acquisitions of business or other foreign direct or indirect investment where there is significant foreign control, to the extent they involve artificial intelligence resources. In April 2026, Innovation, Science and Economic Development Canada launched the Artificial Intelligence Sovereign Compute Infrastructure Program (SCIP), which provides funding for eligible Canadian-owned organisations to build or expand domestic AI compute capacity. The program forms part of the government’s broader Sovereign AI Compute Strategy to reduce reliance on foreign AI infrastructure and support domestic AI development.
The ‘AI for All’ strategy also commits CAD 50 million to expand the Canadian AI Safety Institute to track emerging AI risks, advance technical research, and conduct transparent evaluations of AI models. It proposes creating a Canada Trusted AI Certification program to help Canadians identify trustworthy AI products in the marketplace. The government also intends to work on AI transparency initiatives, including tools such as watermarking AI-generated content.
