Canada’s AI Coordination Gap: Federal Levers Needed to Prevent Market Fragmentation

A key challenge for Canada’s artificial intelligence policy is avoiding jurisdictional fragmentation. As provinces, agencies, and regulators develop procurement rules, public-sector AI requirements, and privacy expectations on separate tracks, Canadian firms – particularly smaller ones – all too predictably face the cost of building to multiple specifications to access what is economically one market.

Interoperability, not necessarily uniformity, should be the objective: A federal baseline for core protections and portable compliance across jurisdictions that avoids market access becoming a patchwork tax on domestic innovation.

Layered federal-provincial regulation or practices become problematic when compliance is non-portable and there is a multiplicity of inconsistent procurement documentation and requirements. Bill C-27 died on the order paper last January, killing both federal privacy modernization and a proposed AI framework.

That has fueled active divergence: Quebec's Law 25 is operational, Alberta has its own privacy legislation and apparatus, and British Columbia is developing its own privacy framework. Each creates different compliance paths that multiply private sector supplier costs without necessarily improving protection.

Public-sector AI procurement rules are similarly proliferating across jurisdictions without coordination. Ontario's Bill 194 (now enacted) creates a cybersecurity framework and contemplates AI system requirements for public-sector entities, including accountability and risk management expectations through future directives and regulations. Similar initiatives will follow in other provinces and, without coordination, they will harden into incompatible regimes.

The potential solution is interoperability – common vocabulary and templates for core artifacts (risk assessments, impact assessments, incident reporting), plus mutual-recognition in procurement – so suppliers can satisfy multi-jurisdictional requirements once rather than recreating equivalent compliance packages province by province.

A federal “AI market access” accord should pursue interoperability: Harmonize what to avoid (harms, minimum governance and transparency expectations for high impact uses, credible redress) while resisting over-prescription of how (system architecture, model choice, technical design). This supports both consumer protection and innovation: Accountability and evidence without micromanaging model development.

A coordination platform exists: The Committee on Internal Trade (CIT), comprised of all Canadian federal, provincial and territorial governments, which was constituted under the Canadian Free Trade Agreement. Problems with such structures are familiar: Coordination tables all too often drift into process unless they have, or are backed by, enforcement leverage that must come from the political level. Nevertheless, we have seen significant progress in the reduction of internal trade barriers under the CIT, as well as from inter-provincial mutual recognition actions, when internal trade gained in prominence on the policy and political agenda.

A similar sense of urgency should pervade the AI ecosystem, giving Canadian AI innovators a better chance to grow.

To function as a meaningful coordination spine rather than an information-exchange forum on this question, the CIT requires:

• An explicit mandate from governments to deliver defined outputs on schedule;
• linkage to federal and provincial procurement rules-makers; and
• a public federal-provincial workplan with measurable deliverables.

Budget leverage means concrete mechanisms: Standardization or mutual recognition of procurement rules and compliance frameworks as eligibility criteria, or federal program funding conditional on provincial alignment with these standards.

Measurable deliverables would include common baseline procurement clauses for high-impact AI systems, a shared incident taxonomy, consistent transparency and redress expectations. The Canadian Digital Regulators Forum (CDRF) – whose members include the Competition Bureau, the CRTC, the copyright board, and the federal privacy commissioner’s office – has demonstrated capacity to produce concrete relevant outputs, including a September paper on synthetic media and related risks across mandates, and should be involved in this exercise.

Finally, the federal “sovereign compute” initiative connects industrial policy to provincial energy and infrastructure decisions. Ottawa is investing $2 billion in domestic AI compute capacity because, without it, Canadian firms train models abroad, researchers depend on foreign capacity, and sensitive applications run outside Canadian jurisdiction.

The federal government should attach transparency and access principles to federally supported compute capacity. Where Ottawa funds that capacity, it should require national-access commitments for qualified Canadian firms and researchers. This disciplined use of federal spending power would prevent local gatekeeping from potentially defeating national objectives.

The core recommendation is consequential: Treat AI “harmonization” as a market-access project. Use the CIT as the federal coordination nucleus but add the incentives and enforcement mechanisms.

Canada’s history shows political sponsorship, budget leverage, measurable deliverables are required. We need to modernize the federal baseline for privacy and AI oversight and make compliance portable through common artifacts. Where Ottawa invests in sovereign compute, insist on transparent, national-access conditions for these computational resources, ensuring Canada's most promising AI entrepreneurs and innovators have access to country-wide capacity.

Canada’s AI market is too small to be fragmented. Let’s accelerate the search for a common regulatory, procurement and access platform into which Canadian innovators and entrepreneurs can lean.

Harvey Naglie has spent 40 years in financial services, and most recently served as a Senior Policy Advisor with the Ontario Ministry of Finance.

To send a comment or leave feedback, email us at blog@cdhowe.org.

The views expressed here are those of the author. The C.D. Howe Institute does not take corporate positions on policy matters.

Want more insights like this? Subscribe to our newsletter for the latest research and expert commentary.