Is AI leading to unemployment? St. Louis Fed weighs in
ST. LOUIS - The July jobs report showed payroll growth slowed to just 35,000 jobs per month from May through July, while the unemployment rate rose to 4.2%. The cause is being debated, and a new analysis from the Federal Reserve Bank of St. Louis points to the rapid adoption of artificial intelligence (AI).
Theoretical AI Use
St. Louis Fed economists looked at theoretical AI exposure, and whether large language models (LLMs) could cut task completion times by half. Jobs with more AI use saw larger unemployment increases between 2022 and 2025. For example, computer and math fields where exposure is around 80% experienced some of the biggest rises in unemployment. Blue-collar and personal service jobs, with far less AI use, saw smaller increases.
AI Exposure and Change in Unemployment across Occupations
Actual AI Adoption
They also looked at real-world adoption data from the Real-Time Population Survey. By late 2024, 23% of U.S. workers reported using AI tools at least once a week. Jobs that used AI most intensively also had the sharpest unemployment gains. Again, technology-related fields stood out, showing significant displacement.
AI Adoption and Change in Unemployment across Occupations
Why It Matters
Unlike past "revolutions" of automation that primarily hit manufacturing and clerical jobs, generative AI targets cognitive tasks performed by knowledge workers—roles once considered relatively secure. While AI may ultimately generate new industries and employment opportunities, this data from the past few years from the St. Louis Fed suggest early signs of the opposite.
The Fed’s View on the Labor Market
Speaking separately at the Peterson Institute, St. Louis Fed President Alberto Musalem said the labor market remains near full employment but warned of a few growing risks. Payroll growth has slowed, unemployment is edging higher, and more vulnerable groups are beginning to feel the strain.
