Canada’s AI Policy Goes Live Through Procurement

Canadian federal AI policy is arriving through procurement rules, audits, and guidance while legislation advances in Parliament. Here is how Ottawa’s evolving AI regulations, from the Algorithmic Impact Assessment to new testing funds, are already changing how companies build and sell AI.

Canada’s AI Policy Goes Live Through Procurement Canada’s federal AI policy is not waiting for the last vote. Ottawa is shaping behaviour through contracts, audits, and guidance that bite today, long before the full legislative package lands. That means developers, founders, and public servants already face practical rules for how artificial intelligence is built, tested, and deployed in government services. The approach is simple, and unusually Canadian, use procurement and risk tools to make AI safer and more accountable now. What is happening, and why now Two tracks are moving in parallel. The first is law. The Artificial Intelligence and Data Act, or AIDA, is the centrepiece of federal legislation bundled with Bill C-27. It proposes a risk-based framework for high-impact AI, the creation of an AI and Data Commissioner, and strong penalties for deceptive or harmful uses. Amendments and consultations have continued, with the details of what counts as high-impact to be defined through regulation after passage. The second track is policy that applies immediately to federal operations. The Treasury Board Secretariat’s Directive on Automated Decision-Making requires departments to co