Canada's AI strategy leaves key safety questions unanswered

Canada's newly-announced AI strategy contains ambitious promises on job growth, adoption, and digital sovereignty—but it's also missing critical details on safety, regulation, and accountability.

The 50-page plan, released Thursday by Artificial Intelligence Minister Evan Solomon, seeks to scale AI adoption from 12 per cent of Canadian businesses to 60 per cent by 2034, and aims to create 250,000 jobs through AI adoption by 2031. But experts and watchdogs point to significant gaps.

There are no clear timelines or key performance indicators for when the strategy's core outcomes should be achieved, and it's unclear who in government is responsible for ensuring delivery. "Canadians, citizens, immigrants, industry, and also public servants, they need to trust the government that they will actually deliver on this and when," said Florian Martin-Bariteau, a law professor at the University of Ottawa.

The strategy mentions upcoming legislation on online harms and privacy but does not say when those bills will be tabled. "We don't know what the scope of that legislation will be, and it will take months for that to work its way through Parliament," said Sara Austin, founder of Children First Canada, adding that parents and children are being asked to "take a leap of faith" while protections remain uncertain.

On job losses, the strategy makes no mention of potential layoffs from AI automation, nor does it detail a plan to support workers affected beyond skills training. This despite Solomon's own acknowledgment that AI "raises hard questions about job security."

Environmental concerns also go unaddressed. While the strategy notes Canada's cold climate and renewable energy advantages for data centre cooling, there are no concrete environmental standards outlined. New research from York University found that Alberta accounts for more than 90 per cent of future AI data centre projects despite relying on a comparatively high-emissions electricity grid.

For Calgary specifically, the AI data centre boom presents both opportunity and risk. The strategy promises doubling Canada's electricity grid capacity by 2050, but without clear environmental oversight or community consultation mandates, local impacts remain uncertain.

The strategy does commit to 850 megawatts of sovereign compute capacity by 2030 through public-private partnerships. But critics argue the document spreads priorities too broadly without a sufficiently clear roadmap for helping Canadian AI companies compete globally.