CountryInitial profile

AI Governance Profile

Canada AI Governance Profile

Initial Observatory profile for Canada AI governance, centered on federal public-sector automated-decision accountability, privacy, AI strategy, the Government AI Register, agentic-AI guidance, AI safety capacity, and enterprise implications.

Last reviewed 2026-09-08
All AI Governance ProfilesAI Workforce Transformation
Canada flag

Profile at a glance

Executive overview

Key signals for Canada AI governance, structured for quick institutional review before the full profile analysis below.

Region

North America

Governance model

Public-sector-led governance

Regulatory maturity

Developing

Enterprise impact

Medium

Public-sector readiness

Advanced

Enforcement maturity

Developing

Major policy and regulatory signals

  • Directive on Automated Decision-Making
  • Algorithmic Impact Assessment
  • AI Strategy for the Federal Public Service 2025–2027
  • Government of Canada AI Register
  • Guide on the Use of Agentic Artificial Intelligence

Enterprise implications

Enterprises should treat Canada as an evolving governance environment rather than a jurisdiction with a single enacted horizontal AI statute.

Last updated

2026-09-08

Key institutions

  • Treasury Board of Canada Secretariat
  • Innovation, Science and Economic Development Canada
  • Canadian AI Safety Institute
  • Office of the Privacy Commissioner of Canada
  • Federal departments implementing automated-decision systems

Watchlist

  • Whether Canada introduces or advances successor federal AI legislation after Bill C-27 / AIDA
  • Any update to the archived AIDA policy architecture or official replacement materials
  • Implementation and research outputs from the Canadian Artificial Intelligence Safety Institute
  • Updates to the Directive on Automated Decision-Making and Algorithmic Impact Assessment tool
  • Federal privacy law reform and Office of the Privacy Commissioner guidance on AI
  • Adoption and evolution of the voluntary code for advanced generative AI systems
  • Implementation milestones for the Canadian Sovereign AI Compute Strategy and AI Compute Access Fund
  • Government AI Register v2 / expanded coverage by end-2026
  • Agentic-AI adoption and governance across federal departments
  • Official sectoral guidance from financial, health, competition, human rights, accessibility, cybersecurity
  • workforce institutions

Source basis

Official source basis

Last reviewed on 2026-09-08.

This profile is grounded in official laws, policy documents, regulator guidance, standards, and public-sector source materials listed below.

Executive summary

Canada’s AI governance architecture is distributed across federal public-sector policy, privacy and sectoral law, safety and research institutions, and broader AI strategy. Its strongest binding operational controls are concentrated in federal administrative decision-making: the Directive on Automated Decision-Making requires risk assessment and proportional safeguards for automated systems used to make or support administrative decisions about clients, supported by a mandatory Algorithmic Impact Assessment. The federal government is also building transparency and capacity through a public AI Register, a 2025–2027 AI Strategy for the Federal Public Service, agentic-AI guidance and the Canadian AI Safety Institute. For private enterprises, compliance exposure remains more fragmented than in the EU or South Korea; suppliers to government should nevertheless expect strong documentation, risk, transparency and assurance expectations.

Governance architecture

Canada’s clearest binding AI governance mechanisms are concentrated in federal public-sector administration. The Directive on Automated Decision-Making and its Algorithmic Impact Assessment establish risk-based controls for covered federal automated decision systems. The federal AI strategy, AI Register, agentic-AI guidance, privacy oversight and Canadian AI Safety Institute add transparency, capability and safety layers. These instruments should not be generalized into a private-sector horizontal AI Act.

Major policies and frameworks

PolicyIssuerYearStatusSummary
Directive on Automated Decision-MakingTreasury Board of Canada Secretariat2019Federal public-sector directive - binding within scopeRequires federal institutions to assess and manage automated decision systems through proportional safeguards, transparency, human involvement, data governance and monitoring.
Algorithmic Impact AssessmentGovernment of Canada / Treasury Board of Canada Secretariat2026Mandatory federal assessment instrument within scopeRisk-assessment tool supporting the Directive on Automated Decision-Making; it does not create a general private-sector AI law.
AI Strategy for the Federal Public Service 2025–2027Government of Canada / Treasury Board of Canada Secretariat2025–2027Federal public-service strategyStrategy for responsible federal AI adoption, workforce capability, service improvement, governance and public trust.
Government of Canada AI RegisterTreasury Board of Canada Secretariat2025–2026Public-sector transparency and implementation instrumentPublic register of federal AI uses and systems; it is an institutional transparency mechanism, not a private-sector AI statute.
Guide on the Use of Agentic Artificial IntelligenceGovernment of Canada2026Official guidance - non-binding outside its administrative contextRisk-based federal guidance for considering agentic AI use, accountability, oversight and operational safeguards.
Canadian Artificial Intelligence Safety InstituteInnovation, Science and Economic Development Canada2026AI safety and evaluation institutionInstitutional capacity for advanced-AI safety research, evaluation, guidance and international coordination; not a general AI regulator.
Bill C-27 / Artificial Intelligence and Data ActParliament of Canada2022Archived proposed legislation - not enactedHistorical proposed federal AI legislation from the 44th Parliament. Bill C-27 ended without enactment and must not be presented as current law.

Policy timeline

2019-04-01

Directive on Automated Decision-Making takes effect

Federal public-sector directive establishes binding administrative-policy requirements for covered automated decision systems.

2020-04-01

Compliance required for systems within scope

Covered federal institutions must comply with the Directive and associated impact-assessment requirements.

2025

Federal Public Service AI Strategy 2025–2027 in operation

Federal strategy links responsible AI adoption with service transformation, governance and workforce capability.

2025-11-28

First Government of Canada AI Register launched

The federal government launched a public register reporting AI uses across institutions.

2026-05-22

Guide on the Use of Agentic AI published

Federal guidance adds an agentic-AI risk and accountability signal without creating a horizontal binding AI law.

2026-09-08

Observatory initial Canada profile review

Initial profile created from official federal, Parliament, regulator, AI safety and public-sector sources.

Enterprise implications

Enterprises should treat Canada as an evolving governance environment rather than a jurisdiction with a single enacted horizontal AI statute. The immediate operational priorities are privacy governance, data stewardship, vendor and model documentation, human oversight, security controls, public-sector procurement readiness, voluntary generative AI commitments where relevant, and monitoring of any renewed federal AI legislation. Organizations selling into or partnering with Canadian federal institutions should pay particular attention to automated decision-making impact assessment expectations and transparency requirements.

Observatory interpretation

Canada’s distinctive signal is institutional layering. The policy system combines binding federal public-sector rules, privacy regulation, open impact-assessment tooling, voluntary generative AI commitments, AI safety institution-building, agentic-AI guidance and compute strategy. AIDA/Bill C-27 remains historical proposed legislation, so practical readiness depends on current administrative, privacy, procurement, safety and accountability controls rather than waiting for an enacted horizontal AI law.

Official resources

ResourceSourceTypeDateLegal forceWhy it matters
Bill C-27, Digital Charter Implementation Act, 2022 — First ReadingParliament of CanadaLaw2022-06-16ProposedProvides the official legislative source for Canada’s proposed federal AI law track, while not establishing binding AI law by itself.
The Artificial Intelligence and Data Act (AIDA) — Companion documentInnovation, Science and Economic Development CanadaGuidance2023DraftImportant for historical interpretation, but the archived status makes legal-status caveats essential.
Directive on Automated Decision-MakingTreasury Board of Canada SecretariatPublic-sector rule2019BindingThis is one of Canada’s clearest binding AI governance mechanisms for federal government use.
Algorithmic Impact Assessment toolGovernment of Canada / Treasury Board of Canada SecretariatPublic-sector rule2026-05-28GuidanceOperationalizes risk classification and mitigation expectations for federal automated decision systems.
Personal Information Protection and Electronic Documents ActDepartment of Justice CanadaLaw2000BindingAI systems involving personal information remain exposed to privacy-law obligations independent of any future AI statute.
Privacy and artificial intelligenceOffice of the Privacy Commissioner of CanadaGuidanceGuidanceShows how Canada’s privacy regulator frames AI-related privacy issues and accountability.
Voluntary Code of Conduct on the Responsible Development and Management of Advanced Generative AI SystemsInnovation, Science and Economic Development CanadaFramework2023VoluntaryProvides a near-term governance signal before any binding horizontal AI statute.
Canadian Artificial Intelligence Safety InstituteInnovation, Science and Economic Development CanadaStrategy2026-06-04Not applicableAdds AI safety institution-building to Canada’s governance architecture.
Guidelines for secure AI system developmentCanadian Centre for Cyber SecurityGuidanceGuidanceConnects AI governance with cybersecurity controls and secure system development.
Canadian Sovereign AI Compute StrategyInnovation, Science and Economic Development CanadaStrategy2026-06-04Not applicableShows that Canada’s AI governance agenda includes compute capacity, data, intellectual property, and infrastructure strategy.
Pan-Canadian Artificial Intelligence StrategyInnovation, Science and Economic Development CanadaStrategy2026-06-04Not applicableProvides the innovation, standards, research, and talent context for Canada’s AI governance profile.
AI Strategy for the Federal Public Service 2025–2027Treasury Board of Canada SecretariatStrategy2025Not applicableShows how federal AI governance is tied to institutional capability and workforce readiness.
Government of Canada AI RegisterTreasury Board of Canada SecretariatPublic-sector rule2025-11-28GuidanceProvides an observable implementation and accountability signal for federal AI deployment.
Guide on the Use of Agentic Artificial IntelligenceGovernment of CanadaGuidance2026-05-22GuidanceAdds a current agentic-AI implementation signal without overstating guidance as binding law.

Update log

2026-06-17: Initial source-backed Canada profile published from official federal, Parliament, regulator, AI safety, privacy, compute, and cyber source materials. AIDA/Bill C-27 is treated as proposed and historical, not binding law.

2026-09-08: Initial published profile created and reviewed from Government of Canada, Treasury Board, ISED and Parliament sources. The Directive on Automated Decision-Making is classified as federal public-sector policy, AIDA/Bill C-27 as historical proposed legislation that did not become law, and the AI strategy, AI Register, agentic-AI guidance and CAISI as strategy, transparency, guidance and institutional-capacity signals.