Goldman Sachs: AI Job Displacement Creates Long-Term Economic Scars
New data reveals AI-displaced workers face longer unemployment and lower re-employment earnings.
1Historical Parallels in Displacement
Goldman Sachs analyzed four decades of labor data to compare current AI-driven layoffs with past technological shifts. Economist Pierfrancesco Mei identified that workers displaced by AI since 2025 face significant re-employment hurdles. These individuals take approximately one month longer to secure new roles compared to other displaced workers, mirroring historical patterns of long-lasting damage.
2The Age-Related Re-employment Gap
Age remains a critical factor in recovery speed following AI-related job losses. Workers aged 35 and older experience a five-week delay in returning to full-time employment, whereas those under 35 face a three-week lag. This trend aligns with broader Labor Department data showing older demographics struggle more significantly with long-term labor market reintegration.
3Quantifying the AI Impact
Challenger, Gray & Christmas reported 120,136 AI-cited job cuts through September, representing 21% of total layoffs. Goldman identified these workers by creating an exposure score based on occupation and industry adoption rates. Data shows that 72.9% of workers aged 25 to 54 were re-employed by January 2026, contrasting sharply with lower rates for older cohorts.
4Earnings and Career Trajectory
Real earnings growth for tech-displaced workers historically lags behind peers, a trend now resurfacing. Only 49% of long-tenured workers displaced between 2023 and 2025 secured roles with equal or higher pay, down from 62% in previous survey periods. Many displaced individuals are forced into roles that offer less stability and lower long-term wage growth potential.
5Skill Transferability and Adaptation
Younger, college-educated, and urban workers often mitigate earnings losses through higher skill transferability. Those with shorter tenures in their previous roles adapted more flexibly, as their skills were less occupation-specific. Conversely, older workers are increasingly moving into new roles that carry similar or even higher levels of AI exposure, potentially repeating the cycle of displacement.
6Entry-Level Hiring Challenges
Stanford’s Digital Economy Lab highlights that AI impact is not limited to layoffs but also affects hiring. Employment for workers aged 22 to 25 in high-exposure occupations is 19% lower than expected. This suggests that AI disruption operates primarily through reduced entry-level opportunities rather than just increasing separations for existing, experienced staff members.
7Methodology and Data Sources
Goldman utilized the Current Population Survey’s Displaced Worker Supplement, analyzing over 15,000 individuals who lost jobs between 2016 and 2025. Since no survey question explicitly asks about AI, researchers constructed a proprietary score. This methodology allowed them to isolate the first wave of AI-displaced workers, noting that displacement trends were statistically insignificant before 2023.
8Divergent Economic Perspectives
While Goldman highlights significant scarring, other institutions remain cautious. The Yale Budget Lab and various researchers report no widespread, economy-wide job displacement directly linked to AI metrics yet. These conflicting findings suggest that while specific sectors face acute disruption, the broader labor market impact remains a subject of ongoing, complex economic debate.