Half of Canadians use AI at work but lack training & confidence

A recent survey has found that nearly half of Canadians are using artificial intelligence tools at work, but only a small fraction have received formal training, revealing a significant skills and confidence gap across the workforce.

The study, conducted by Growclass in partnership with Angus Reid, surveyed over 1,000 Canadian adults and illustrates both regional and demographic divides in AI adoption, confidence, and concerns about the future of work. While 42% of Canadians stated they are currently using AI in their jobs, just 12% have had any formal instruction in how to use these tools effectively.

Confidence levels

Despite close to half the workforce employing AI in their daily activities, only 33% of respondents expressed confidence in their ability to use these tools efficiently. This discrepancy highlights a gap between growing workplace expectations and the practical training provided to employees.

The survey also indicates gender disparities in AI confidence. Thirty-six percent of men reported feeling confident in their AI skills compared to just 30% of women. Moreover, men were slightly more likely to have received formal AI training, with 13% indicating so versus 11% of women. Job security concerns also presented a gendered trend, with 29% of women expressing concern about AI replacing their job within the next five years, compared to 22% of men. Overall, 26% of Canadians indicated they were worried about job displacement from AI in the near future.

Regional and sector differences

Interest in upskilling and AI training is not evenly spread across Canada. Respondents from Ontario and Quebec demonstrated the highest demand, with 71% of surveyed individuals from both provinces expressing willingness to undertake AI training offered by their employers. In contrast, just 56% of Atlantic Canadians showed the same level of interest.

A breakdown by profession suggests that HR professionals are the most frequent AI users (81%) and lead in training received (33%). Despite this, they are also most concerned (62%) about being replaced by AI within five years. On the other hand, professionals working in non-profits reported the highest confidence in AI use at work (92%), closely followed by communications and marketing (90%), and digital technology sectors (85%).

Responsibility and barriers

The majority of those surveyed (46%) believe their employer should be responsible for AI training, compared to 20% who think it should be up to the individual. Smaller proportions assigned this responsibility to education systems (7%) or government (4%), while 16% were unsure who should provide such training. In total, 68% would participate in job-offered training if given the chance.

Among respondents not interested in employer-provided AI training, the predominant reasons cited were lack of trust or concerns about bias (68%), fears that AI might decrease critical thinking (67%), ethical or privacy issues (61%), and the cost associated with AI tools (13%).

Growclass launches certification

"AI is already central to how we work, but without proper training, it risks leaving people behind, particularly for women and underrepresented communities," said Sarah Stockdale, CEO and Founder of Growclass. "Our program is designed to close that gap by giving people the practical skills, confidence and value alignment they need to use AI intentionally, not fear it."

To address these gaps, Growclass has announced the launch of its AI Marketing and Strategy Certification. The initiative is structured as a six-week programme targeting marketing professionals, particularly women and underrepresented groups. The certification offers hands-on experience with AI tools, guidance from industry experts, live workshops, and a roadmap for integrating AI into day-to-day marketing practices. Graduates also gain ongoing access to course materials and membership in the Growclass alumni community.

The survey was carried out in both English and French, with a sample deemed broadly representative of the Canadian adult population. The margin of error associated with the findings is +/-3.1 percentage points, 19 times out of 20.