Skills gap persists

Companies are buying artificial intelligence faster than they are training people to use it, even as the hardest-to-automate work still depends on human judgment. (mckinsey.com) McKinsey said in January 2025 that 92% of companies planned to increase artificial intelligence investment over the next three years, but only 1% of leaders said their companies were “mature” in deployment. The firm said the main barrier was leadership, not worker readiness. (mckinsey.com) In a November 2025 report, McKinsey said more than 70% of skills sought in job postings are used in both automatable and non-automatable work. Its Skill Change Index found digital and information-processing skills were most exposed to automation, while assisting-and-caring skills were among the least exposed. (mckinsey.com) That split helps explain why employers still need managers, negotiators and team leaders even as software takes over more routine tasks. McKinsey said future work will be organized around people, software agents and robots working in the same workflow, not around full human replacement. (mckinsey.com) Canada shows the same tension in the data. Statistics Canada said on June 16, 2025 that 12.2% of businesses had used artificial intelligence to produce goods or deliver services in the previous 12 months, up from 6.1% a year earlier. (statcan.gc.ca) Artificial intelligence use was highest in information and cultural industries at 35.6%, professional, scientific and technical services at 31.7%, and finance and insurance at 30.6%, according to Statistics Canada. Text analytics was the most common application, used by 35.7% of firms that had adopted artificial intelligence. (statcan.gc.ca) KPMG said in June 2025 that 89% of 250 Canadian business leaders were investing in upskilling, but 53% said their companies still did not invest enough in training, workshops or continuous learning. The same poll found 56% said they lacked the internal talent and resources to implement technology effectively, and 63% said employees were not using new technology effectively enough to deliver full returns. (kpmg.com) A separate April 2, 2026 analysis from the Canadian Federation of Independent Business found nearly 45% of Canadian businesses now use generative artificial intelligence in their operations. Firms that invest in artificial intelligence were 5.4 percentage points more likely to invest in employee training, and nearly 8 in 10 said they planned to maintain or increase training spending in 2026. (cfib-fcei.ca) The numbers point to a mixed picture: adoption is spreading, but training is uneven and often late. That leaves the most valuable work clustered around tasks machines still struggle with, including coaching, persuasion, conflict management and decisions that depend on trust. (mckinsey.com) McKinsey put the choice plainly in its 2025 and 2025 reports: companies can bolt artificial intelligence onto old jobs, or redesign workflows so software handles routine steps while people handle the parts that require judgment. The firms that close the training gap first are more likely to find out whether that partnership model actually pays off. (mckinsey.com)