Canada’s artificial intelligence minister says deploying AI systems needs to be regulated, but Evan Solomon declined Sunday to say what form that regulation will take, according to a CP24 report published Sept. 27.
Asked whether Ottawa plans legislation to restrict how AI is deployed, Solomon said: “I’m not going to say ‘yes, this is what we’re doing.'” He pointed instead to measures already in place or under way, including a national AI strategy, data privacy legislation and social media safety rules, and said the government has finished a public consultation on AI transparency. “Safety is our number one concern,” he said, according to CP24.
Solomon also acknowledged concern about AI agents that could improve themselves without authorization or behave deceptively. “We’re tracking that closely,” he said. He told CP24 he has raised with G7 and G20 counterparts the idea of slowing the release of new AI models and creating an international technology stability board. No timeline for either was given.
The comments leave Canada in an awkward spot. The previous government’s Artificial Intelligence and Data Act, part of Bill C-27, died when Parliament was prorogued in January 2025, and BetaKit reported in October 2025 that Solomon said he would not revive it. At the time he promised what he called “light, tight, right” rules that would not “choke off” innovators, and said he would table legislation covering data transfers, children’s sensitive data and deepfakes, according to BetaKit. Almost a year later, no such bill has been described in detail, and experts quoted by BetaKit have warned that Canada trails peers such as the European Union, which already has its AI Act in force.
What Solomon can point to is money and diplomacy. At the All In conference in Montreal on Sept. 16, he announced $300 million in Canadian and German funding for LawZero, the safety-focused nonprofit led by Mila founder Yoshua Bengio, according to BNN Bloomberg. The same event produced a Canada-Germany research partnership between Mila and the German Research Center for Artificial Intelligence, and Solomon framed the broader push as a sovereignty question. “If we don’t build it here, we’ve got to buy it from someone else. If we don’t innovate here, we’ve got to rent it from someone else,” he said.
Industry leaders have their own view on oversight. Aidan Gomez, chief executive of Toronto-based Cohere, told the same conference that “a multinational body is a much better solution than just a bunch of islands trying to regulate this themselves,” BNN Bloomberg reported. That position lines up with Solomon’s interest in an international stability board, but it also raises a practical question for Canadian companies: which rules apply in the meantime.
For Canadian businesses, the uncertainty is the story. Firms building or buying AI tools, from banks to health-care providers to startups, have no federal statute telling them what transparency, testing or incident-reporting duties may come. Those operating in Europe must already meet the EU’s requirements, while those selling only at home are working from general privacy law and voluntary codes. Without a bill, compliance planning is guesswork.
Solomon’s answer Sunday suggests the government is weighing its options rather than preparing an imminent tabling. Whether that patience reassures innovators or frustrates those who want firm safeguards will depend on how quickly the promised detail arrives. The minister has said safety comes first, and the next test is whether that priority shows up in a draft law Canadians and companies can actually read.
