StatCan’s first AI workforce survey finds AI still has to prove its place

A majority of Canadian workers not using generative AI at work said the tools don’t apply to their jobs. In Statistics Canada’s first worker-focused TechStat survey, released today, 56% of non-users gave that answer. 

A training session may not change that answer. You can teach someone to fish, but they still need a reason to go fishing.

The survey, based on a Labour Force Survey supplement collected in March 2026, found 35.9% of workers aged 15 to 69 had used generative AI tools at work in the past 12 months. 

Management occupations led at 75.1%, followed by natural and applied sciences at 67.5%. By industry, professional, scientific and technical services came in at 65.6%, with finance, insurance, real estate, rental, and leasing at 59.2%. 

Accommodation and food services had the lowest industry rate at 16.3%. By occupation, trades, transport and equipment operators came in lowest at 14.7%.

Nearly two-thirds of users applied the tools to some but not most tasks. A quarter used them for almost none, and only 11.6% used them broadly.

The survey also shows the limit of counting users. Someone who applied generative AI to almost none of their tasks still counts in the 35.9% use rate.

Beyond the 56% who said the tools don’t apply, another 21.6% of non-users said they had no interest. Only 5.8% cited a lack of skills or knowledge. 

Security, privacy, environmental, or ethical concerns came in at 9.8%, rising to 16.9% among public sector employees. Another 5% said workplace policies limited their use.

The age data complicates the assumption that the youngest workers lead AI adoption at work. 

In occupations like engineering, teaching and health care, where AI has more potential to complement the work, workers aged 25 to 54 used generative AI at 56.9%. The 15-to-24 group came in at 39.1%, though the younger cohort in degree-required fields is small.

Among users, seven in 10 reported using the tools daily or a few times a week. StatCan will ask again in September 2026. 

When those results arrive, we might have a better sense of how many workers simply hadn’t found a useful application for generative AI, and how many were right that their jobs had little use for it.

Final shots

• A worker can count as an AI user after applying it to almost none of their tasks. Ask what changed in the work before treating the user count as proof of value.
• Before ordering more training, ask non-users where generative AI could help in their jobs. Some jobs may not have a useful answer.
• Workers aged 25 to 54 led generative AI use in occupations where it has the most potential to complement the work. Employers should not assume the youngest workers are already the furthest ahead.