Regulatory Landscape
Canada was an early mover in AI governance, launching the world’s first national AI strategy in 2017 and implementing automated decision-making requirements for federal agencies in 2019. However, comprehensive federal AI legislation remains elusive after AIDA (the Artificial Intelligence and Data Act) died with the dissolution of Parliament in early 2025.
Canada’s approach reflects its position as a significant AI research hub—home to pioneers like Yoshua Bengio, Geoffrey Hinton, and major research labs—balanced against the desire to maintain competitiveness with the US.
Key Laws
AIDA, introduced as part of Bill C-27, would have established requirements for “high-impact” AI systems including mandatory risk assessments, measures to address bias, transparency obligations, and the creation of an AI and Data Commissioner. The bill underwent extensive committee review but did not pass before Parliament dissolved.
The Directive on Automated Decision-Making, applying to federal government systems, requires algorithmic impact assessments, peer review for high-impact systems, and meaningful explanations for affected individuals. It serves as a model for what federal AI regulation might look like.
The voluntary code of conduct for generative AI, signed by major companies operating in Canada, covers safety testing, transparency, content labeling, and research collaboration.
Enforcement
Currently, enforcement relies on existing authorities: the Privacy Commissioner for data-related AI issues, the Canadian Human Rights Commission for discrimination, and sector-specific regulators. The proposed AI and Data Commissioner would have had dedicated AI enforcement powers.
Business Impact
The regulatory uncertainty following AIDA’s failure to pass creates a gap for businesses seeking clear compliance frameworks. Many Canadian AI companies default to EU AI Act compliance given its extraterritorial reach and the likelihood of serving EU customers. Canada’s strong AI research ecosystem continues to attract investment despite regulatory ambiguity.