Carney Government Prepares Reintroduction of Artificial Intelligence and Data Act After Years of Stalled Progress

The Carney government is preparing to reintroduce the Artificial Intelligence and Data Act, the federal framework for artificial intelligence regulation that was first tabled by the previous Liberal government and that died on the order paper before the 2025 federal election, with officials at Innovation, Science and Economic Development Canada working on a reworked version of the legislation that is intended to align more closely with European Union and other allied approaches and to address concerns that had been raised by industry, civil society, and international observers about the original draft. The reintroduction, expected in the coming months, would represent the most significant federal regulatory action on artificial intelligence in Canadian history and would establish Canada's first comprehensive horizontal AI regulatory framework.

The regulatory context

The Artificial Intelligence and Data Act, commonly referred to as AIDA, was originally tabled as part of Bill C-27 in 2022. The bill was the federal government's response to the growing recognition that artificial intelligence systems were being deployed across Canadian society in ways that produced significant implications for individuals, for institutions, and for the broader public interest, and that existing regulatory frameworks were not adequate to address those implications.

The original AIDA was structured around the concept of high-impact AI systems, which would be subject to specific regulatory requirements including risk assessments, monitoring, and transparency obligations. The bill was the subject of significant engagement from industry, civil society, academic, and international stakeholders across its parliamentary consideration. The bill ultimately died on the order paper before the 2025 federal election without having received Royal Assent.

The international regulatory environment has continued to evolve significantly in the period since AIDA was first tabled. The European Union's AI Act has come into force and has begun to be implemented. The United Kingdom has been developing its own AI regulatory framework, with a generally lighter-touch approach than the European Union but with continuing engagement on high-risk applications. The United States has implemented executive-branch frameworks under both the Biden and second Trump administrations, with significant variation in approach across the two administrations.

What the reworked AIDA may include

The reintroduced AIDA is expected to maintain the core structural approach of the original bill, with regulatory obligations applied to high-impact AI systems while leaving lower-impact applications to existing legal frameworks. The detail of how high-impact systems are defined, what specific obligations apply to them, and how enforcement is structured are expected to be significantly revised relative to the original draft.

Federal officials have signalled that the reworked bill will engage more fully with the European Union's AI Act framework, both because the European framework provides the most developed international model and because alignment with the European approach has commercial value for Canadian companies operating in European markets. The European AI Act's risk-based classification system and its specific obligations for general-purpose AI systems are areas where the reworked AIDA is expected to engage more directly.

The reintroduced bill is also expected to address concerns that had been raised about the original draft's structure, including questions about the level of detail provided in primary legislation versus regulations, about the scope of ministerial discretion, and about the relationship between the AI regulatory framework and existing privacy, consumer protection, and human rights frameworks.

Industry engagement

The Canadian AI industry has been engaged in continuing consultation with the federal government on the reworked legislation. The Canadian Council of Innovators, the Information Technology Association of Canada, and individual companies including Cohere and others have been engaging at varying levels of intensity. Industry views on the reworked legislation have been varied, with some emphasising the importance of regulatory certainty and others emphasising the importance of preserving Canadian competitive position relative to lighter-touch regulatory environments.

The Canadian AI research community, including the Vector Institute, Mila, and the Alberta Machine Intelligence Institute, has similarly been engaged in the consultation process. The research community's perspective has emphasised both the importance of effective regulation and the need to ensure that the regulatory environment supports continued Canadian leadership in AI research and commercialisation.

Civil society organisations, including the Centre for Digital Rights, the Canadian Civil Liberties Association, OpenMedia, and several others, have engaged on the privacy, human rights, and democratic accountability dimensions of AI regulation. The civil society perspective has emphasised the importance of strong regulatory protections, particularly for vulnerable populations and for democratic institutions, and has been pressing for stronger AIDA provisions than were included in the original draft.

The international alignment question

The international alignment of Canadian AI regulation is a significant consideration in the reworked legislation. Canadian commercial AI companies operate increasingly in international markets, and the regulatory environment in those markets shapes the operational requirements for Canadian companies. Alignment with the European Union's AI Act is expected to be a particularly important objective given the European market's significance and the EU framework's status as the most developed international model.

The Canadian Sovereign AI Compute Strategy and related industrial policy initiatives have been positioning Canada as a participant in the broader international conversation about AI governance. Effective regulation that aligns with international approaches is part of the broader strategic positioning, given that international customers including governments increasingly look for sovereign AI options that operate within familiar regulatory frameworks.

The continuing political environment in the United States, including the second Trump administration's general posture of regulatory rollback, produces complications for the international alignment question. American AI regulation has been less developed than European or Canadian frameworks, and the regulatory direction under the current American administration has been generally toward less rather than more regulation. The Canadian alignment with European approaches rather than with American approaches is one element of the broader Canadian strategic posture toward international policy alignment.

Privacy and data protection

The AIDA framework intersects with the broader Canadian privacy and data protection environment. The Personal Information Protection and Electronic Documents Act, the Quebec privacy regime, and provincial public-sector privacy frameworks all interact with AI regulation. The reworked AIDA is expected to address the relationship between AI regulation and privacy regulation, with the goal of producing a coherent overall framework rather than overlapping or conflicting requirements.

The federal Office of the Privacy Commissioner has been engaged in the broader conversation about AI regulation. The Privacy Commissioner has issued guidance on AI applications and has been working with the federal government on the design of regulatory frameworks. The continuing role of the Privacy Commissioner in the AI regulatory environment is one of the structural questions that the reworked AIDA will address.

Implementation and enforcement

The implementation and enforcement structure of AIDA will be a significant element of the reintroduced legislation. The original draft proposed an Artificial Intelligence and Data Commissioner within Innovation, Science and Economic Development Canada, with rule-making and enforcement authority. Concerns had been raised about the structure, including questions about the independence of the commissioner and about the relationship to other regulatory authorities.

The reworked legislation is expected to address those concerns. Specific structural arrangements for the regulatory authority, for the relationship to other federal regulators, for the relationship to provincial regulators, and for the relationship to international regulatory authorities will all be addressed in the reintroduced bill.

The enforcement powers under AIDA, including investigation, audit, and penalty authorities, will similarly be elements of the reworked framework. The European Union's AI Act provides one model for enforcement structure, although Canadian implementation will need to be adapted to the specific Canadian constitutional and regulatory environment.

Provincial dimension

Provincial governments have continuing roles in AI regulation through their authority over privacy, consumer protection, professional regulation, and provincial public-sector applications. Quebec has been particularly active in establishing its own AI-specific regulatory framework, with the province having implemented requirements for AI applications used in certain provincial contexts. Other provinces have been working on their own approaches at varying paces.

The federal-provincial coordination on AI regulation is a continuing area of work. Effective implementation of AIDA depends in part on appropriate coordination with provincial regulatory frameworks. The Carney government has indicated it intends to work closely with provincial counterparts on the implementation of the reworked AIDA.

What's next

The reintroduction of AIDA is expected in the coming months as part of the Carney government's broader legislative agenda. The bill will go through standard parliamentary consideration, including committee study, witness testimony, amendment proposals, and ultimately parliamentary votes. The Liberal majority means the bill can advance through Parliament if the government chooses to prioritise its consideration.

Industry, civil society, academic, and provincial engagement with the reintroduced bill is expected to be significant. The various perspectives that engaged with the original AIDA will engage with the reworked version, with the substantive changes from the original draft producing both new opportunities for engagement and new questions to be addressed.

For Canadians, the AIDA reintroduction represents one of the more significant pieces of forthcoming federal legislation in the technology and economic regulation space. Whether the reworked bill produces an effective Canadian AI regulatory framework, whether it aligns appropriately with international approaches, and whether it supports continued Canadian competitiveness in the broader AI economy are the central questions that will be answered through the parliamentary process and the subsequent implementation period.