KPMG reveals the workforce gap quietly draining employer investment
Canadian employers are redesigning jobs around AI — and workers aren't keeping up.
A KPMG Canada survey of 306 executives found that only three percent of Canadian organizations have achieved measurable returns on their AI investments, even as 70 percent say the technology is already delivering meaningful business value.
The primary obstacle, respondents said, is a workforce skills gap.
The gap is widening partly because employees are pushing back.
According to the KPMG survey, 31 percent of Canadian workers resist agentic AI — systems that operate independently with minimal human oversight — compared with 16 percent globally.
Over half of resistant employees cite trust and ethical concerns, while nearly 40 percent point to job security fears or a lack of confidence in their own AI capabilities.
The stakes are rising quickly.
Two-thirds of executives surveyed say they are moving toward a fully integrated AI-human workforce, with 77 percent already using agents for tasks such as knowledge sharing between departments.
Within two to three years, 39 percent expect AI agents to lead project management for teams outright.
That shift is already hitting the talent pipeline.
According to KPMG, 59 percent of respondents say agents have changed how they hire entry-level workers, and 63 percent say the same about experienced talent.
Organizations are now prioritizing creative thinking (46 percent), problem-solving (44 percent), and adaptability (43 percent) — with 65 percent of business leaders saying social and interpersonal capabilities now outweigh technical expertise.
Performance management is changing alongside hiring.
Respondents anticipate AI collaboration competencies being built into reviews and role requirements (39 percent), greater emphasis on distinctly human skills over tasks now handled by AI (39 percent), and redefined promotion criteria centred on AI literacy and effective agent delegation (36 percent).
KPMG Canada partner Lewis Curley said AI has moved beyond background support and is now shaping how organisations think about talent.
To unlock value, he said, everyone needs to know how to allocate work, assess quality, and "give effective feedback."
Curley said training alone will not close the gap.
Organizations also need to show employees precisely how working with agents will produce results — otherwise, he said, the investment stalls before it pays off.
Leading firms are responding by establishing clear protocols for agent use, building sandbox environments for hands-on practice, and developing role-specific delegation guidelines.
