JPMorgan Warns of AI-Driven Jobless Recovery as Optimists Dispute Outlook

JPMorgan economists have raised alarms over the potential for artificial intelligence to trigger a “jobless recovery” in the U.S. labor market, particularly for white-collar knowledge workers. In a recent analysis, senior U.S. economist Murat Tasci highlighted that the widespread adoption of AI tools could displace roles requiring non-routine cognitive tasks—such as those in finance, law, engineering, and corporate administration. These occupations, which accounted for 45% of U.S. employment in 2025, have historically shown resilience during economic downturns but now face heightened risks as AI automates tasks previously reserved for human expertise title1^[1]. Tasci noted that non-routine cognitive workers now constitute a larger share of the unemployed than non-routine manual workers, a shift he described as “ominous” and indicative of rising structural unemployment risks title2^[2].

The warning stems from evolving patterns in the labor market. Over the past four decades, routine jobs—both cognitive (e.g., sales, office work) and manual (e.g., construction, production)—have seen prolonged recoveries after recessions, with employment in these sectors still below pre-2008 levels. However, AI’s ability to replicate cognitive tasks threatens to extend this trend to high-skill roles. For instance, generative AI tools are already reducing demand for entry-level professionals in finance, legal services, and tech, where tasks like contract review, data analysis, and content generation are increasingly automated title3^[3]. JPMorgan data reveals that unemployment among college graduates has risen to 5.8%, with majors in AI-exposed fields like computer engineering and graphic design disproportionately affected title4^[4].

The economic implications of such displacement could be severe. A jobless recovery, where GDP and corporate profits rebound but employment lags, risks prolonged unemployment for white-collar workers and could dampen consumer spending. Tasci warned that anemic recovery prospects for non-routine cognitive occupations might mirror past jobless recoveries seen in routine sectors, exacerbating wage polarization and geographic disparities as urban hubs with high concentrations of knowledge workers face sharper job losses title5^[5]. Meanwhile, Goldman Sachs Research estimates that AI could displace 6–7% of U.S. jobs, though it cautions that these effects are likely temporary as new roles emerge .

Early data suggests AI’s impact is already materializing. The Bureau of Labor Statistics and JPMorgan analysis show that non-routine cognitive workers now account for a larger share of unemployment than routine workers, a reversal of historical trends title6^[6]. Sectors like cloud computing, web search, and computer systems design have seen employment plateau since 2022, coinciding with the rise of large language models title4^[4]. Additionally, the Federal Reserve Bank of St. Louis found a 0.47 correlation between AI exposure and rising unemployment in occupations such as computer and mathematical roles, reinforcing concerns about displacement title9^[8].

Not all experts share JPMorgan’s pessimism. Tech investor David Sacks, the White House’s AI and crypto czar, argues that AI will not replace workers but rather augment productivity, creating a competitive edge for those who master the technology. He emphasized that AI still requires human oversight for context, prompting, and validation, rendering apocalyptic job loss scenarios overhyped title1^[1]. Similarly, Goldman Sachs notes that historical technological shifts have led to net job creation, with 60% of today’s occupations nonexistent in 1940 .

As AI adoption accelerates, policymakers and corporations face mounting pressure to address potential disruptions. JPMorgan urges proactive measures, including upskilling programs, career pathway redesign, and enhanced social safety nets, to mitigate structural unemployment risks. Meanwhile, the Federal Reserve’s recent data underscores the need for vigilance: even as the overall unemployment rate remains near historic lows at 4.2%, sector-specific trends suggest a more nuanced labor market transition title9^[8].