Home Artificial Intelligence (AI)More than half of Canadian workers now use AI, but most say employer training falls short

More than half of Canadian workers now use AI, but most say employer training falls short

by Todd Humber
A+A-
Reset

More than half of Canadian employees now use generative artificial intelligence at work, but the majority say their employer’s training programs are inadequate and they lack clarity on workplace AI policies, according to new research from KPMG.

The firm’s 2025 Generative AI Adoption Index found that 51 per cent of Canadian workers use AI tools at work, up from 46 per cent in 2024 and 22 per cent in 2023. However, the pace of adoption has slowed compared to the previous year’s growth.

Only 48 per cent of employees said the training their employer provided has been helpful, while 83 per cent said they want or need to learn how to use generative AI tools more effectively. Nearly half — 46 per cent — said they worry about losing their job if they can’t keep up with AI advancements.

Technology overload hampering adoption

The survey revealed significant challenges in implementation. Thirty-six per cent of employees said they received training but haven’t started using AI tools because they are too busy to implement new work processes. Another 37 per cent said they started using AI after training but stopped because it was too overwhelming.

“Employees were using AI in their personal lives and were often ahead of their companies to begin with, but now, employers are catching up and need to engage their workforce more broadly to drive adoption,” said Megan Jones, a partner in the People and Change practice at KPMG in Canada.

Jones said the slowdown in adoption growth could reflect growing pains as organizations move from experimentation to concrete integration of AI technology.

Concerns about accuracy limit use

Concerns about AI-generated errors are holding workers back. Fifty-eight per cent of employees said they are extremely or very concerned about hallucinations or inaccuracies in AI responses, and 46 per cent said this concern prevents them from using the tools at work.

“It’s encouraging to see how eager employees are to use AI but it’s apparent they are scared of losing their jobs, overwhelmed by the pace of change and feel underprepared for what’s ahead,” said Jones. “If this continues, organizations risk losing the productivity gains AI is meant to provide.”

Policy confusion widespread

The survey found significant gaps in workplace AI governance. Only 29 per cent of employees said their employer has a comprehensive AI policy, up from 18 per cent in 2024. Forty per cent said they don’t know what AI controls are in place at their organization.

Nearly half of workers — 49 per cent — said using AI should be a mandatory policy at all Canadian organizations.

“Canadian organizations have an opportunity to turn employee interest into measurable impact. But to achieve this, they can’t leave employees guessing at the expectations around AI use,” said Davin Gnanapragasam, chief technology officer and tax partner at KPMG in Canada.

Source: 2025 KPMG in Canada Generative Al Adoption Index (CNW Group/KPMG LLP)

Productivity gains reported

Despite the challenges, 79 per cent of employees said generative AI has improved their productivity. Among those who use AI regularly, 73 per cent said they use the tools daily to a few times per week, up from 64 per cent in 2024.

Fifty-one per cent of employees said they redeploy the time saved by AI to higher-value work, up from 45 per cent in 2024.

The most common uses include:

  • Research: 62 per cent
  • Idea generation: 58 per cent
  • Summarizing public information: 37 per cent

Training must be role-specific

Gnanapragasam said organizations need to provide personalized training that teaches employees how to identify specific use cases relevant to their roles.

“AI success isn’t about being the most tech-savvy, it’s about being the most adaptable. Training should empower employees to understand how to strategically apply AI to their day, so it frees up time for more value-added work, rather than a one-size-fits-all solution,” he said.

The survey was conducted from Aug. 15 to 29, 2025, and included 2,239 Canadian employees aged 18 and older. The margin of error is plus or minus three percentage points, with a 95 per cent confidence level.

Related Posts

Leave a Comment