So Far, AI Is Reshaping More Than Cutting the Workforce

AI is more likely to reallocate work, reshape career paths and raise productivity than cause widespread mass unemployment.

• Since 1940, 60% of workers have moved into jobs that did not exist then, while technology-driven new positions accounted for 85% of employment growth.
• Firms with the highest generative-AI spending per employee increased white-collar employment by 10.2% more than peers during the first two years after adoption, although their faster underlying growth limits the causal interpretation.
• Research covering more than 65 million U.S. workers across about 280,000 firms links AI exposure to weaker junior employment relative to senior employment, mainly through slower hiring rather than layoffs; Stanford payroll data show a 19% relative decline for early-career workers in exposed occupations.
• CEO surveys indicate a shift from headcount reduction toward reskilling, retention and role redesign, with nearly half of executives focusing on reskilling and a similar share combining human and AI capabilities.
• The OECD estimates that AI could add 0.4%–1.3% annually to labor-productivity growth in highly exposed G7 economies through 2034, versus 0.2%–0.8% in less-exposed countries, supporting higher growth and a more disinflationary environment.

Takeaway: Investors should favor businesses and industries that redeploy talent and combine AI with human expertise, while recognizing that early-career disruption, weaker consumption and uncertainty about the speed of labor-market adjustment remain important risks.