Future of work

Research from Statistics Canada examines how generative AI has affected Canadian employment since late 2022 by comparing occupations with different potential exposures and complementarity to AI.  

The research finds that overall employment increased across all occupation types, with no clear evidence that jobs more exposed to AI declined faster than others. However, the report emphasizes that it is difficult to isolate the impact of AI because recent labour market trends also reflect post-pandemic adjustments, demographic shifts, immigration growth, and trade tensions. This framing suggests that early fears of massive AI-driven job losses are not yet supported by aggregate data. However, distributional effects across groups may be emerging beneath the surface.  

The report groups jobs into three distinct categories based on how occupations could interact with AI technologies relative to the median occupation:  

• High-exposure and low-complementarity (HE-LC) occupations are expected to involve a greater number of tasks that may be more susceptible to replacement by AI.  

• High-exposure and high-complementarity (HE-HC) jobs have a larger share of tasks that are more likely to be augmented by AI  than to be replaced.  

• Low-exposure occupations are not expected to rely significantly on AI.  

While the research found no evidence that growth in HE-LC jobs or coding-intensive occupations slowed compared to the rest of the labour market, this trisection of jobs by AI exposure reveals the same underlying trends that have been observed across the labour market. Regardless of how a worker’s occupation interacts with AI, employment grew faster in larger establishments (those with 500+ workers), and younger and less-educated workers experienced weaker employment growth than older and more educated workers. Employment among workers aged 15–29 grew only modestly compared with stronger gains among those aged 30–49, indicating stagnation at the entry level even during a period of overall employment expansion. This aligns with some of LMIC’s findings on the struggles of youth and new grads in today’s labour market.  

Furthermore, a regression analysis also shows that after the introduction of generative AI in late 2022, industries with higher shares of HE-LC employment (controlling for pre-pandemic and telework trends) did not show a significant slowdown in employment growth or weekly earnings.  

Overall, the evidence points to gradual distributional effects rather than to abrupt disruption. As with earlier technological waves (such as the introduction of computers and increases in automation and digitization), the main pattern appears to be one of unequal gains, with higher-skilled workers advancing more quickly while youth and entry-level workers lag. This pattern suggests that generative AI is less a sudden shock and more a continuation of long-run technological change, with varied impacts across different groups of workers.