62% of Canadian workers say AI is eroding their skills, IBM finds — while HR barely uses AI itself

IBM published the largest study yet of how AI is landing inside the workplace today — and the headline is a warning, not a celebration. The IBM Institute for Business Value, working with Oxford Economics, surveyed 1,500 CHROs and 8,800 employees across 21 geographies. In Canada, 62% of employees say they worry AI is eroding their skills — and 69% of those say the erosion has already begun, with critical thinking the skill cited as declining most often.

The timing is awkward for the profession sounding the alarm. While 68% of Canadian CHROs name the ability to supervise, validate and override AI outputs as the workforce's most essential skill, 63% of Canadian organizations make limited or no use of AI inside the HR function itself. The people being asked to lead the transformation are, by their own ratings, barely using the technology.

The headline numbers#

The two surveys ran from April to June 2026 — 75 Canadian executives among the CHRO sample, 452 Canadian employees among the workers — and they sketch a workforce split by seniority. Canadian CHROs identify critical thinking (55%) and human judgement (49%) as among the workforce's most important capabilities in the AI era. But only 29% of Canadian employees rank judgement as important, while 54% of employees and 55% of CHROs agree that critical thinking and problem framing matter most. The disconnect is sharpest where it matters: the people running AI strategy say judgement is the skill of the decade; most employees haven't clocked that yet.

Who gets blamed when AI fails#

Accountability is the study's ugliest finding. 39% of Canadian employees report that when something goes wrong with AI, the blame falls on them — while 32% of Canadian CHROs admit unclear accountability complicates AI deployment. The companies that get this right see measurable payoffs: organizations that clearly define which workflows are human-led, AI-assisted or AI-executed report an 18% reduction in risk and a 20% improvement in quality. Yet 37% of Canadian organizations don't involve the CHRO at all when AI strategy is being defined.

There is also a participation gap with real consequences for safety. Where the CHRO at least shares responsibility for deciding which decisions stay human-led, 76% of employees say they feel safe to question or override AI recommendations — compared with 43% where HR is merely advisory. The difference between a workforce that pushes back on a bad AI output and one that rubber-stamps it may come down to whether HR has a seat at the strategy table.

The invisible work AI creates#

87% of Canadian CHROs believe AI adoption creates "invisible" work for employees — validating recommendations, fixing mistakes, providing context, managing exceptions. Half of employees agree: 51% of Canadian employees say AI increases their workload or that the extra work goes unrecognized. 45% of Canadian CHROs say the productivity gains AI does deliver are primarily reinvested into innovation or reskilling, which means workers absorb the toil while the gains flow elsewhere.

HR's own lag#

The most pointed finding is reserved for the function itself. Canadian CHROs rate their own teams' capabilities as low across the board: 23% for AI literacy, 20% for AI performance measurement, 24% for change management around AI adoption. Globally, organizations with mature HR AI capability — governance, architecture, measurement — are nearly twice as likely to report a higher number of positive business KPIs. "As AI becomes more deeply embedded in how work gets done, organizations have an opportunity to strengthen the human capabilities that matter most," said Brandon-Jo Mouna, Director of Human Resources at IBM Canada. "Critical thinking, judgement and the ability to challenge AI-generated outputs will become increasingly important as employees work alongside AI."

What to watch#

Two things will determine whether this study ages as a snapshot or a turning point. First, whether CHROs move from the sidelines: the 76%-vs-43% split on employees feeling safe to override AI is the single most actionable number in the report, and it argues for HR involvement in AI strategy as a safety control, not a courtesy. Second, whether the "invisible work" gets counted — 87% of CHROs acknowledge it exists, and as long as validating AI output remains uncredited labor, skills erosion will keep feeling like a pay cut in disguise. The full study is published by the IBM Institute for Business Value; the Canadian findings were released via CNW today.