Goldman Sachs Study Finds AI Is Measurably Squeezing Labor Markets in Narrow Sectors
Summary
- • Goldman Sachs finds AI measurably slowing hiring in specific sectors since mid-2022.
- • Call-center employment fell 39% below trend in the U.S., 33% in Canada, 27% in Germany.
- • Entry-level workers bear the heaviest AI-related headcount growth drag across economies.
- • AI adoption averages 15–20% across major developed markets; France and U.S. lead adoption.
Details
Goldman analyzes 800+ occupations for AI exposure
Goldman Sachs analyzed more than 800 occupations to map AI exposure levels to employment outcomes across major developed economies.
U.S. call-center jobs 39% below trend
U.S. call-center employment stood 39% below historical trend; Canada 33% below, Germany 27% below — among the steepest declines in any sector.
Entry-level headcount drag per 10% AI exposure
A 10% AI occupational exposure is linked to a 0.1 pp annual headcount drag overall; entry-level roles face a 0.2 pp drag in the U.S. and over 0.6 pp in Australia.
Hardest-hit sectors identified
Call centers, software publishing, management consulting, and advertising services show employment furthest below long-run historical trends across developed economies.
AI adoption at 15–20% across developed markets
AI adoption averaged 15–20% across major developed economies; France, U.S., Netherlands, and UK are leading adopters according to Goldman's analysis.
AI labor impact remains narrow but deepening
Goldman concluded that AI-related hiring pressures are real and measurable but are currently limited to a narrow set of industries and worker categories, with broader labor markets relatively insulated.
Goldman Sachs research published August 19, 2026 via CNBC. Details compiled via Grok live web research (CNBC, IFA Online).
What This Means
Goldman Sachs' analysis provides the most detailed institutional view yet of how AI is reshaping labor markets, confirming that displacement is real but concentrated. Call centers and entry-level roles in knowledge work are bearing the brunt, while broader labor markets remain relatively insulated. This research adds significant empirical weight to the policy conversation around AI and jobs, with granular data pointing to specific sectors that may need targeted workforce transition support. The narrow but deepening impact in exposed occupations suggests effects will widen as AI adoption spreads beyond its current 15–20% average.