Experts question Anthropic report on AI's impact on work

In September 2026, Anthropic released an economic impact report outlining an extreme scenario regarding the large-scale adoption of AI. The projection suggested that by 2030, the U.S. could experience a significant surge in GDP, while the share of national income going to capital would rise from about 40% to nearly 55%, leaving workers with a shrinking slice of the economic pie. Under this scenario, widespread job displacement by AI could drive the U.S. unemployment rate close to 12%. For comparison, the typical U.S. unemployment rate during normal economic times hovers around 4% to 5%, whereas it peaked at roughly 25% during the Great Depression in the 1930s. By October 23, 2026, this report and its underlying macroeconomic logic drew criticism and debate among economists. Macroeconomic researchers such as Tom Orlik pointed out that, based on economic history, a simultaneous surge in both growth and unemployment is unrealistic, questioning who would purchase the increased output if everyone were unemployed. Even though the realization of this extreme scenario remains controversial, the risk of AI causing massive unemployment continues to draw attention. According to calculations by related economic research teams based on the proportion of tasks in various occupations where AI can play a substantive role, about 26% of the U.S. labor force—equivalent to roughly 45 million workers, including those in the financial sector—faces potential disruption.