Canada now faces a very familiar problem: weak productivity growth at the exact moment a powerful new workplace tool is spreading fast.
A 2024 report from the Bank of Canada says labour productivity growth declined over the past three decades, and a 2025 economic survey warns that trade tensions could make that weakness even more costly.
Against that backdrop, the new adoption and productivity update from the St. Louis Fed deserves attention in the U.S.
By August 2025, it found 54.6 per cent of respondents aged 18 to 64 had used generative AI, up from 44.6 per cent a year earlier, and it estimates the technology may already have lifted U.S. labour productivity by as much as 1.3 per cent.
Canadians are not sitting this out.
A national survey of 1,500 adults from Toronto Metropolitan University found that 66 per cent of Canadians had tried generative AI by early 2025, less than three years after ChatGPT’s debut.
The catch, is intensity and understanding. The same study found only about 30 per cent were using these tools daily or weekly for work, study, or leisure, while concerns about privacy, job loss, and critical thinking remained high.
That mix of curiosity and unease feels distinctly Canadian: the public is moving faster than the rules, and faster than many employers.
Business uptake is where the gap shows.
According to last year’s Canadian Survey on Business Conditions, 12.2 per cent of businesses reported using AI to produce goods or deliver services in the second quarter of 2025, up from 6.1 per cent a year earlier.
In those data, the leaders were information and cultural industries at 35.6 per cent, professional, scientific and technical services at 31.7 per cent, and finance and insurance at 30.6 per cent.
Yet the national picture is still hesitant. In the third-quarter business outlook, 14.5 per cent of firms planned to adopt AI over the next year, while 66.7 per cent had no such plans and 18.9 per cent were uncertain.
The case for moving faster is no longer theoretical. In a writing experiment with 453 college-educated professionals, access to ChatGPT cut completion time by 40 per cent and improved quality by 18 per cent.
In a customer support study by Microsoft covering 5,000 agents, a generative AI assistant raised productivity by 14 per cent on average, with the biggest gains among less experienced workers. Those are task-level gains, not an automatic national boom, but they help explain why the U.S. data are starting to show broader movement.
Canada’s early labour-market evidence is more measured than the hype.
A recent labor-market study found employment generally grew from November 2022 to December 2025 regardless of how exposed occupations were to AI, although younger and less educated workers saw weaker gains.
The same research notes that businesses using AI did not report a surge in job cuts, with the share saying AI reduced employment holding at about six per cent.
A Bank of Canada paper on the AI productivity paradox makes a similar point: micro-level gains are real, but diffusion lags and organizational adjustment mean the macro payoff takes time.
That is the Canadian choice in 2026. We can treat generative AI as another imported gadget that workers use quietly while institutions hesitate, or we can treat it as a chance to attack a productivity problem that has dogged Canada for years.
The lesson from the U.S. report is not that Canada has fallen behind forever. It is that broad adoption arrives before most leaders feel ready.
The winners will be the firms, schools, and public institutions that move from scattered experiments to better training, clearer guardrails, and redesigned jobs before the rest of the country is forced to catch up.
