Anthropic Economist Says AI Has Not Lifted 4.2% U.S. Unemployment Despite 10%-20% Job-Loss Warning
Updated
Updated · Fortune · Jul 24
Anthropic Economist Says AI Has Not Lifted 4.2% U.S. Unemployment Despite 10%-20% Job-Loss Warning
1 articles · Updated · Fortune · Jul 24
Summary
Peter McCrory, Anthropic’s head of economics, said 18 months of internal research and updated BLS data show no material AI-driven rise in U.S. unemployment, directly challenging CEO Dario Amodei’s near-term job-loss forecasts.
June unemployment held at 4.2%, job openings were roughly in line with the number of unemployed workers, and prime-age employment remained near multi-decade highs—conditions McCrory said do not fit a broad AI displacement story.
McCrory added that workers in highly AI-exposed occupations have not seen relative unemployment deterioration versus less-exposed roles, arguing Claude’s "jagged" capabilities still require human oversight and work more as augmentation than replacement.
The essay still flags weaker hiring for young workers in AI-exposed roles and slower long-term growth projected through 2034 for jobs such as technical writing, data entry and customer support.
That leaves Anthropic straddling two views: Amodei has warned AI could erase half of entry-level white-collar jobs and push unemployment to 10%-20%, while McCrory says any severe disruption has not yet appeared in aggregate labor data.