Canada joins a growing list of countries with AI registries
by Piyush Mathur
With the launch of its Artificial Intelligence (AI) register in late November 2025, the North American country of Canada has joined a growing list of national polities that have already established such portals. Canada’s launch was announced on the website of the Treasury Board of Canada Secretariat (TBCS) on November 28, 2025.
The TBCS notes that the register displays the government-verified records of ‘42 institutions and features over 400 systems where AI is currently being explored, developed, implemented, or deployed’ in Canada.
The other national polities that have been ahead of Canada in this regard are the Netherlands, the United Kingdom (UK), Norway, and China; additionally, agency-wide and region-wide AI registries have existed for sometime at several places around the world, including the cities of Amsterdam (the capital of the Netherlands) and Helsinki (the capital of Finland). Moreover, at least eight civil-society AI registers—including one from Canada itself—antedate the Canadian government’s initiative.
These registries are not structured the same, have the same scope, or work on the same principles; broadly, however, they aim to systematize and make transparent how AI and algorithmic systems are being used by government agencies.
The registry is yet another milestone in Canada’s AI policy
The Government of Canada AI Register, a public-facing catalogue, tracks Canada’s federal AI systems used in service delivery and administrative decision-making—and is accessible through an open-data portal. In its current form, it is labelled a Minimum Viable Product, implying that it is ‘an early version with only basic features and content that is used to gather feedback.’
While the TBCS had also issued a Directive on Automated Decision Making on March 4, 2019, this AI Register is not described as a part or a renaming thereof—but as a new initiative that delivers a key instrument promised in the July 9, 2025 AI Strategy for the Federal Public Service 2025-2027.
Both these directives, however, are part of Canada’s unfolding policy on technological accountability and transparency in governance. The July 2025 AI strategy statement includes a mention of the 2019 directive, aside from references to similar directives regarding data strategy, digital talent, digital services, and digital standards.
The government has also put together on its website a veritable history of the initiatives it has taken on using and managing AI in government. Going all the way back to a reference titled ‘Drafting of the AI whitepaper (October, 2016 – October, 2017)’, this history includes the mandatory 6-monthly updates to the Directive on Automated Decision-Making (2019)—and a brief description of Canada’s leadership in putting together the G7 AI Network (GAIN), founded on September 10, 2025.
The AI Register is promising, but it won’t blunt the ongoing criticism of Canada’s AI governance initiatives
Canada’s AI Register reinforces the country’s ongoing shift from ad hoc transparency efforts toward formal, structured disclosure regimes; it also aligns with an emergent global movement toward algorithmic transparency and mandatory risk assessment. The initiative can only benefit citizens in regard to digital governance. However, Canada still has some way to go before it could satisfy its civil society’s concerns whose priorities are being addressed via the government’s push toward the deployment of AI in governance, and the associated theme of accountability in AI’s usage in the private sector.
Within 5 days of the TBCS’ announcement of this AI Register, for instance, Renee Black, the Chief Executive Officer (CEO) of the Vancouver-based non-profit organisation GoodBot, complained in a TechPolicy.Press article that ‘public-interest researchers—experts in privacy, digital rights, Big Tech accountability, and online harms—are systematically sidelined’ in Canada’s AI policy articulations. Previously, the country had a failed Artificial Intelligence and Data Act—which means that it currently has no federal legal code specific to the governance of AI.
Among Black’s list of recommendations for Canada is her plea to slow down the federal consultation process for the AI strategy—a plea that has been made by several individuals and groups previously, including by the British Columbia Civil Liberties Association.
Meanwhile, resentment appears to be growing among Canada’s non-White minorities, especially Blacks, regarding their exclusion from Canada’s AI Strategy Task Force. Writing for Policy Options, the Montréal-based Institute for Research on Public Policy’s (IRPP) newsletter, a law professor recently expressed his shock that this fairly large task force of 27 included not a single Black member in its original lineup even though Blacks constitute ‘the very community most adversely impacted by AI’.
Concluding words
While Canada is far more systematic and transparent than the vast majority of countries around the world in how it is going forward with its AI governance, it must grasp the fact that claiming a leadership role globally would require it to heed its mounting domestic criticisms. That Canada desires such a leadership role is obvious from its explicit declarations as well as the nature of its references to the Organisation for Economic Co-operation and Development (OECD) and G7 in its AI governance announcements.
Canada must also keep in mind that China has been trying its best to look good to the world regarding its AI transparency—and it even has a global AI governance proposal, the World Artificial Intelligence Cooperation Organization (WAICO), to advance.
Piyush Mathur, Ph.D., is the author of Technological Forms and Ecological Communication: A Theoretical Heuristic (Lexington Books / Bloomsbury, 2017); for several years, he was an adjunct scholar with McMaster University’s Institute on Globalization and the Human Condition.