A study by economists Nir Jaimovich and Henry E. Siu found that 88% of routine-occupation job losses clustered around U.S. recessions, informing debate over the labor risks of an AI investment shock.
Their analysis of U.S. employment since the mid-1980s placed those losses in the 12 months surrounding recessions dated by the National Bureau of Economic Research. The finding describes past downturns; it does not establish that AI will produce the same pattern.
In an Oct. 8 analysis, Michael Cembalest, chairman of market and investment strategy for J.P. Morgan Asset & Wealth Management, calculated that 42 AI-related stocks accounted for roughly 60% to 80% of S&P 500 returns, earnings growth and capital spending since January 2024. The range covers those three measures, rather than a single share of market gains.
The Bank of England’s Financial Policy Committee warned in September that a sharper fall in AI-related assets could spread more widely as debt financing for AI investment grows. After AI and semiconductor shares declined in July, the committee found no spillover into core markets or signs of wider financial stress.
A separate study of U.S. job advertisements found larger increases in skill requirements where the Great Recession hit harder. The difference persisted through at least 2015. Its authors said the evidence was consistent with production shifting toward routine-biased technologies and greater demand for workers whose skills complement them.
The International Labour Organization (ILO) estimated in May 2025 that about one in four workers worldwide held a job with some exposure to generative AI. It expected most jobs in that group to change rather than disappear. In a separate study covering 135 countries, the ILO estimated exposure at 30% to 32% of employment in high-income countries, compared with 10% to 15% in low-income countries.
The ILO identified women, younger workers and clerical staff among groups with greater exposure. The International Monetary Fund found that job openings requiring AI skills carried higher wages. It also associated the spread of AI-linked skills with lower employment in occupations highly exposed to AI where the technology offers workers little support.
The Budget Lab at Yale found no clear evidence of AI-related disruption in U.S. labor data through August 2026 in its Sept. 15 update. The Bank of England is scheduled to publish its next Financial Stability Report and Financial Policy Committee record on Nov. 26.
