Updated July 30, 2026, 8:20 a.m. Pacific. Future trends should be checked against the next Statistics Canada survey. Generative AI is now part of Canadian work, but the pattern is less dramatic than the idea that every job has suddenly been rewritten. Statistics Canada’s first regular Labour Force Survey supplement on the subject found that 35.9% of workers aged 15 to 69 had used generative AI in their main job or business during the previous 12 months.
The more revealing number is depth of use. Among users, 63.5% applied the tools to some, but not most, tasks. Only 11.6% used them for most or almost all tasks. AI is common enough to matter, yet it is still usually an assistant rather than an autopilot for an entire job.
Awareness and practical adoption are also different things. While 93.4% of workers had heard of generative AI, 51.5% were familiar with how it could apply to their current work, including 15.0% who felt very familiar. Another 37.4% knew the tools but did not think they applied to their work. Knowing the name of a chatbot is not the same as having a useful, safe workflow for it.
| What the survey measured | Share | How to read it |
|---|---|---|
| Used generative AI at work in the past year | 35.9% | Any use; it does not mean daily use or that AI performs the whole job |
| Users applying it to some tasks | 63.5% | Assistance with selected writing, research, analysis or coding tasks is the mainstream pattern |
| Users applying it to most or almost all tasks | 11.6% | Deep use remains a minority and cannot be inferred from the overall adoption rate |
| Users working with it daily | 31.4% | Another 38.3% used it a few times per week, so regular use takes several forms |
Occupation makes a large difference. Use reached 75.1% in management and 67.5% in natural and applied sciences, compared with 14.7% among trades, transport and equipment operators and 17.0% in natural resources and agriculture. By industry, professional, scientific and technical services stood at 65.6%, finance, insurance, real estate, rental and leasing at 59.2%, and educational services at 53.0%; accommodation and food services stood at 16.3%. These gaps largely reflect how much text, data and digitizable work a job contains. They are not a ranking of which workers are more advanced.
Statistics Canada also grouped occupations by potential AI exposure and by whether AI is more likely to complement their tasks. Generative AI use was 53.8% in high-exposure, high-complementarity occupations, 45.9% in high-exposure, low-complementarity occupations and 14.2% in low-exposure occupations. The index describes the potential for tasks to change; it is not a forecast of job losses. Doctors, teachers and engineers may see some tasks augmented, while parts of retail, office support, software development or accounting may be more susceptible to replacement. Actual outcomes still depend on economic, institutional and technological choices.
Non-use is not simply a skills problem. Among the 64.1% who had not used generative AI in their main work over the past year, 56.0% said it did not apply to their work, 21.6% had no interest and 9.8% cited security, privacy, environmental or ethical concerns. Only 5.8% identified a lack of skills or knowledge, while 5.0% cited an employer or organizational policy. For employers, the practical sequence is to identify an appropriate task first, then provide a secure tool, clear boundaries and training.
For a Canadian job search, it is more useful to describe a concrete AI-assisted workflow than to put “familiar with ChatGPT” on a résumé. Explain how the tool helped draft, test code or compare information, then describe the human verification, privacy boundary and final decision. Our Canada life hub and Canadian job-search guide provide broader context. For the policy side, continue with our Canada AI transparency consultation explainer. This is workplace data reporting, so no affiliate link was inserted in the article body.
The central figures come from Statistics Canada’s July 30 release, checked against the March 2026 questionnaire, its June workplace AI profile and the federal generative AI at work guidance. This was a sample survey, not a census. Its population excluded people living on reserves, full-time members of the regular Armed Forces and people living in institutions, and no percentage should be read as an individual worker’s probability of replacement.