Canada’s ‘AI for All’ strategy signals plan to roll out AI safeguards through various laws
The federal government released its highly-anticipated national artificial intelligence strategy on Thursday, outlining a legislative plan that will likely create AI safeguards through multiple laws rather than a single overarching bill, bring privacy standards in line with those in the European Union and California, and introduce more protections for children, according to lawyers who work with the technology sector.
They add that the strategy appears to strike a compromise between strengthening protections for individuals against AI on the one hand, and developing a regulatory landscape that won’t deter AI companies from establishing themselves in Canada on the other.
The legislative plan largely focuses on addressing the “more pressing consequential harms” related to AI use, like the proliferation of deepfakes or misinformation, says Max Jarvie, a partner at Davies Ward Phillips & Vineberg LLP who advises clients on privacy, data protection, and AI technologies. He notes that the plan says much less about the responsible development of AI tools.
“There’s a legislative gap that I think has been deliberately chosen because of the overarching importance of ensuring that Canada stays at the forefront… of AI development,” Jarvie says.
“It is a bit of a trade-off because it brings the risk of allowing for the development of such technology in a more unshackled way, but [requires addressing] any harms that come up in a post-facto fashion.”
Ariel Laver, a partner at Fasken who advises clients on acquiring, transferring, and developing technology, noted that many members of Canada’s tech industry were critical when the federal government tabled its last attempt to establish a comprehensive regulatory framework for AI, the Artificial Intelligence and Data Act (AIDA). AIDA died in 2025 when Prime Minister Justice Trudeau prorogued Parliament.
“The concern that I’ve heard raised from players in the industry was that AIDA was actually potentially even more restrictive than the EU AI Act,” Laver says. He adds that as a “middle power” with a smaller economy than the US, the European Union, and other jurisdictions, “it would be difficult for Canada to have the most restrictive legislation and keep companies that are developing AI products within Canada.”
The fact that the strategy unveiled on Thursday doesn’t seem to involve plans to reintroduce another bill like AIDA will likely be “positively received by the industry,” Laver says.
Named “AI for All,” the federal government’s AI strategy revolves around six pillars that cumulatively aim to better protect Canadians against AI-related risks, improve AI literacy, increase businesses’ adoption of AI tools, and reduce Canada’s reliance on foreign AI infrastructure by building out its own.
The first pillar focuses on safety, and involves introducing consumer privacy legislation to strengthen privacy rights, protect children, and increase Canadians’ control over their personal data. The pillar also involves plans to introduce online safety laws, and ongoing review of the federal Privacy Act, which governs how the federal government handles Canadians’ personal data, so that it aligns with international standards.
The strategy noted that relative to other countries, Canada ranks low in AI literacy and training, and that half of Canadians regard AI as a threat to humanity. It did not provide details about specific legislative proposals or timelines.
Canada currently lacks a straightforward regulatory framework for AI. For protection against AI-related issues, Canadians rely on a patchwork of rules outlined across data privacy and human rights laws, the Canadian Charter of Rights and Freedoms, province-specific rules, and court decisions.
Thursday’s AI strategy appears to indicate that this patchwork approach will not change, Jarvie says, adding he would be surprised if the federal government introduces another comprehensive AI bill like AIDA. AIDA was one of the components of Bill C-27, an omnibus bill that also aimed to modernize Canada’s data privacy rules.
“I don’t want to suggest that there will be no legislation that touches on AI issues, but I don’t get the impression that there’s going to be a standalone law on AI in the form of [AIDA] as it was under Bill C-27,” he says. “I think the strategy as it stands suggests that there will be ways to address concerns arising from AI that will be put into other pieces of legislation.”
Laver had a similar takeaway.
“There’s signals here that rather than potentially launching another version of AIDA, it seems as though the federal government is signaling that they’re really focusing on filling in gaps that exist in legislation,” Laver says. That includes updating Canada’s data privacy rules to better align with the EU’s General Data Protection Regulation and California’s Consumer Privacy Act, but also covering bases that aren’t addressed by current laws, like concerns about individuals or companies not being transparent about their AI use, AI in social media, and protecting children’s personal information, he says.
The strategy noted that deepfakes are currently being used as a form of sexual violence, particularly against women and children, and that one of the federal government’s priorities is safeguarding children’s online information from exploitation and harm.
Overall, Jarvie said he’s gratified the federal government is addressing AI-related issues, many of which are “now long-standing issues from a legislative perspective,” like AI-related online harms and surveillance pricing.
“We’ve really been waiting for some kind of update from the federal government on all of these issues for quite some time,” he says.
