Statistics Canada did not find broad labor collapse
Statistics Canada's February 2026 study on generative AI and employment says overall employment generally grew from November 2022 to December 2025 regardless of whether occupations were classified as more or less exposed to AI. The same paper says coding-intensive jobs grew at a similar rate as other jobs.
Statistics Canada explicitly says it is still hard to separate any AI effect from post-pandemic adjustments, demographic shifts, and trade tensions.
The youth labor market was already weak enough to make new pressure matter
The same Statistics Canada study says younger employees aged 15 to 29 saw much weaker employment growth than workers aged 30 to 49 over the same period. It also says gains in coding-intensive jobs were concentrated among workers aged 30 to 49, while the number of coding professionals younger than 30 stagnated.
Outside the AI paper, the labor-market backdrop is already rough. Statistics Canada's youth labour reporting says the unemployment rate for people aged 15 to 24 hit 14.7 percent in September 2025 and was still 14.1 percent in February 2026. Returning students aged 15 to 24 faced 17.9 percent unemployment in the summer of 2025, the highest summer rate since 2009 excluding 2020.
Adoption is moving in one direction even if layoffs are not yet the headline
Statistics Canada's business-conditions survey says 12.2 percent of businesses had used AI over the prior 12 months by the second quarter of 2025, and 14.5 percent planned AI use over the next year by the third quarter of 2025. The same release says most AI-planning firms expected no employment change, but 12.2 percent expected employment decreases while only 7.3 percent expected increases.
The business survey puts rising adoption beside a minority, but larger, group of AI-planning firms expecting employment decreases than increases.
Hiring may show the first damage before unemployment does
Anthropic's March 2026 labour-market report, using U.S. Data, found no systematic increase in unemployment for highly exposed workers so far. But it also reported tentative evidence that job finding among workers aged 22 to 25 had slowed in the most exposed occupations, with a roughly 14 percent drop relative to 2022.
The Canadian and Anthropic records both put early-career workers in the risk lane before any broad unemployment shock appears in headline labor data.
The entry lane is where the pressure appears first
If AI pressure lands first on job finding rather than total employment, the workers most exposed may be the ones trying to get onto the ladder rather than the ones already established. Under-30 coding stagnation matters even without a dramatic topline unemployment event.
A weak entry lane can reshape careers long before a national labor market shows obvious collapse. Delayed starts, weaker first jobs, slower skill accumulation, and fewer openings in exposed occupations can all matter even while aggregate employment still looks stable.
