Canadian CEO poll finds 46% expect AI to support growth with fewer workers

Growth without a growing payroll is becoming part of Canadian executives’ AI outlook. In EY-Parthenon’s Canada CEO Outlook, published October 28, 46% of 50 surveyed business leaders said many companies could meet long-term growth targets with fewer workers. Yet a much larger share expects AI’s nearer-term impact to fall on how people work.

The productivity measure is output, not cuts

The mechanism executives describe is doing more without adding people or capital. Shane Dunn, managing partner at EY-Parthenon Canada, said an overwhelming majority of Canadian CEOs report measurable returns from AI investments. Their main productivity measure, he said, is higher output without additional headcount or capital—not necessarily a reduction in the staff already employed.

That distinction separates two workforce outcomes. Producing more with the same team is different from meeting growth targets with a smaller team. Dunn’s account describes the first as a measure of returns; the survey’s 46% finding captures executives’ belief in the second over the long term. Neither figure is a count of jobs eliminated by AI.

Earlier Canadian polling gives more detail on the changes executives anticipated. In a separate Spring 2026 Canada CEO Outlook, conducted in March and April, 80% of 50 executives planned to increase AI spending that year. None planned to cut it. Spending was divided roughly evenly among efficiency automation, restructuring complex workflows, improving quality and business-model innovation.

Workforce changes named in the spring poll

• 46% ranked teaching existing employees new or more advanced skills among their top two expected workforce impacts.
• 34% cited redesigning roles to combine human and AI capabilities.
• 20% cited reducing hiring in certain roles—a different measure from laying off existing employees.

46%Long-term growth with fewer workers

This share of respondents believes many companies can reach long-term growth targets with smaller workforces.

82%Greater impact on how work is organized

This share expects roles, skills and workplace organization to change more than workforce size over the next three years.

A window into large-company thinking

The latest Canadian responses were collected in August and September. EY said most participants lead publicly traded companies with annual sales above US$1 billion. Nearly a quarter reported annual revenue between US$10 billion and US$49.9 billion. Their businesses span sectors including consumer goods, finance and health care.

Those details matter to the interpretation: this is a small sample of leaders at large businesses, not a survey of Canadian workers or a tally of employment changes. And the 46% question concerns what respondents believe many companies can achieve. It should not be read as 46% announcing cuts at their own organizations.

EY’s global findings offer context for the productivity argument. In its October 1 release, 50% of surveyed CEOs identified AI as the largest contributor to productivity gains over the previous year. But 23% said AI-driven gains were often absorbed before reaching bottom-line results. That survey covered 1,200 CEOs across 21 countries, rather than just the Canadian group.

Changing tasks does not mean eliminating jobs

Viet Vu, manager of economic research at Toronto Metropolitan University’s Dais think-tank, told CBC News that researchers do not yet see a negative labour-market impact they can attribute to AI, including among young people. He located the fear in what might happen next, while recommending that office and white-collar workers prepare to learn new skills.

TD Economics senior economist Thomas Feltmate made a related distinction in an October report. He predicted gradual labour-market adjustment rather than a sudden shock, describing an AI jobs crisis as a risk scenario, not the base case. His reasoning rests on the difference between tasks and jobs: a job combines tasks requiring different skills, judgment and expertise. AI’s ability to transform some of those tasks does not, by itself, establish that the whole position will disappear.