On January 6, 2025, Justin Trudeau announced that he would resign as Canada’s Prime Minister and on March 9, 2025, the Liberal Party of Canada chose Mark Carney as its new leader. With Parliament now prorogued, and an election scheduled to take place on April 28, 2025, Canada’s proposed national artificial intelligence legislation, the Artificial Intelligence and Data Act (“AIDA”), part of Bill C-27, has died on the order paper. Any future AI law will require either re-introduction of AIDA by the Liberal Party or a re-write by whichever party forms the next government.
In this bulletin, we consider the public positions of Canada’s major political parties and their MPs relating to AI, based on numerous rounds of parliamentary debate and committee hearings since Bill C-27’s introduction. In sum, there appears to be a consensus among every major political party that Canada needs a comprehensive and effective national AI regulatory framework. That said, the best approach to legislating in an area of rapid technological change, the breadth of the regulation that would accompany the legislation and the specific enforcement mechanisms to employ are all hotly debated.
Liberal Party of Canada
The Liberal Party’s position is well established and is reflected in AIDA, which was tabled by the Minister of Innovation, Science and Industry (the “Minister”). As stated in AIDA, the legislation aimed to achieve two primary objectives: (i) regulating international and interprovincial trade and commerce in AI systems by establishing common requirements across Canada for their design, development, and use; and (ii) prohibiting conduct related to AI systems that could cause serious harm to individuals or their interests. In particular:
- Primary Goals of AI Legislation – AIDA was drafted with the intention of ensuring responsible development and deployment of AI systems, including:
- establishing guidelines regarding unlawful conduct and the need for risk and impact assessment;
- ensuring transparency; and
- mitigating harm and discrimination bias, particularly in high-impact AI technologies.
- Deferring Rulemaking to Future Regulators – The proposed legislation and subsequent amendments grant the Governor in Council and the Minister the authority to establish future regulations. These regulations will, among other things, establish the mechanics for Canada’s approach to AI legislation including the appropriate enforcement mechanisms, fix record keeping rules, and will define the parameters for what constitutes material harm. The Liberal government views the deferral of regulation as the best way to allow Canada’s AI laws to evolve alongside AI innovations.
- Specificity and Scope – On November 28, 2023, the government proposed amendments to AIDA to, among other things, focus the definition of “artificial intelligence systems” so that it clearly applies to both general-purpose systems and high-impact systems, identify several initial classes of “high-impact systems” (the first version of AIDA left this entirely to future regulation); and shift certain oversight and administrative responsibilities from the Minister to a person who will be designated as the “Artificial Intelligence and Data Commissioner.” As will be set out below, these changes were in response to criticisms from the other political parties. While a new Liberal government might make adjustments in any future artificial intelligence legislation, it is fairly safe to assume AIDA will be used as a guide.
Conservative Party of Canada
As of this writing, polls indicate that the Liberal Party will succeed in the upcoming election, though as recently as March 18, 2025 the Conservative Party was in the lead. Although the Conservative Party under Pierre Poilievre has yet to publish a formal position on AI regulation, speeches and comments from MPs suggest that the Conversative Party would make several changes to the overall approach. For example, several Conservative MPs, led by shadow Minister of Innovation, Science and Industry, Rick Perkins, have argued during the bill’s second reading and in committee that:
- Specificity and Scope – AIDA is a “massive overreach”, and the promise of future consultation for developing an AI regulatory framework is “too little, too late” for effective regulation.
- Responsibility for Rulemaking – AIDA’s broad language would “pass a law that defines no goals or oversight” and effectively consolidated much of the power into a single minister, granting an anti-democratic overreach in authority to oversee critical areas for innovation through “massive financial and jail penalties” for breaches in law and regulation not yet defined.
- Commissioner Independence – Without clear boundaries in the proposed legislation there are concerns that the Minister could delegate broad regulatory powers to an AI and Data Commissioner who reports directly to the Minister which could create a conflict of interest undermining the balance between economic development and protecting citizens.
- Privacy and AI – The marriage of artificial intelligence and facial recognition presents a significant risk to the privacy and other rights of Canadians, but was not addressed by AIDA.
- Regulatory Clarity – The lack of a clear regulatory framework would push innovation out of Canada and into countries with clearer laws. A clearer and more narrowly focused regulatory framework at the outset would better promote innovation and improve Canada’s position in the rapidly grown AI-driven economy.
While the parliamentary debates surrounding AIDA preceded the November 28, 2023 amendments, many of the Conservative Party’s criticisms weren’t fully addressed by the proposed amendments. It appears that a Conservative government would want to narrow the application of any future AI legislation and/or clearly define, at the outset, the how the law would be implemented and enforced through accompanying regulation. This is reflected in the numerous rounds of parliamentary debates and committee hearings since Bill C-27’s introduction.
New Democratic Party (NDP)
The NDP, though not the leading party, supported the Liberal minority on a number of issues. Cracks in the relationship are evident in the NDP’s views on AIDA. Like the Conservatives, the NDP have not clearly articulated their stance on AI law and regulation. With that said, NDP MP Brian Masse has been a vocal advocate forAI innovation and regulation, voicing several concerns during the AIDA’s reading and committee consideration, some of which overlap with the issues raised by the Conservatives, including:
- AI Oversight Committee – Whether it would be prudent to create a House and Senate Committee (such as with defence) to oversee artificial intelligence to provide comprehensive and long-term oversight of AI in Canada.
- Regulatory Clarity on the Commissioner’s Power – Establishing a clear initial regulatory framework to define and empower the role of an AI and Data Commissioner is crucial for effective oversight and regulation. A properly defined framework at the outset would help avoid the challenges currently faced by the Privacy Commissioner and the Competition Bureau.
- Robust and Adaptable Enforcement Power – In order to ensure that the AI and Data Commissioner will be effective and can evolve with the growth of AI, he or she will need robust tools and enforcement powers. There is a concern that without real strength and a sufficient budget, the Commissioner will be unable to address the complexities that arise with the expansion of AI.
Bloc Québécois (Bloc)
The province of Quebec has taken a proactive stance on ethical AI governance and Bloc MPs Sébastien Lemire and Jean-Denis Garon have been active in parliamentary debate and committee meetings. For example, the Bloc has pointed out that:
- Provincial Consultation and Cultural Protections – The amended meaning of “high-impact systems” overlooks models that threaten the integrity of minority cultures and cultural diversity. With globalization making culture more homogenized, the regulation of AI should not be limited to strictly regulatory and technological issue but should also consider cultural issues that require consultation with the provinces.
- Safeguards on AI and Law Enforcement – With recent concerns over the use of facial recognition technologies, the use of artificial intelligence could possibly place the presumption of innocence at risk. Future AI legislation should address these risks.
- Stronger Protections for Copyright – AIDA does not address any issues of AI and copyright. Artists have a fundamental right to control what happens to their works. There should be some level of transparency and protection for artist to strengthen the protection of creators’ copyrights, particularly in a context where cultural and linguistic preservation is a priority.
Conclusion
The impending election will likely result in a fresh look for Canadian AI legislation. Understanding the positions of the major Canadian political parties provides important insights into how each of the major parties might legislate and regulate if they form a government, or part of a governing coalition. While this bulletin highlights some of the differences among the major Canadian political parties, there is, fortunately, a consensus on the underlying urgency. Establishing a nationwide AI legislative framework should be a top priority for any new government.
Special thanks to our articling student Daniel Dai for contributing to this update.
