🔍 Read the full analysis: Why Goldman Sachs Compares AI Layoff Scars With Past Tech Shocks on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
A Goldman Sachs client note, reported by Fortune, estimates that workers likely displaced by AI since 2025 took about a month longer to find new jobs and were roughly 10 percentage points more likely to remain unemployed a year later than other laid-off workers. Goldman compares those early results with its historical research on technology-driven job losses, which found slower reemployment and lasting wage penalties. The AI-specific estimates are not independently confirmed because the note is not public and the attribution method is indirect.
Workers likely displaced by AI since 2025 have taken longer to find new jobs and have been more likely to remain unemployed a year after losing work, according to findings from a Goldman Sachs client note reported by Fortune. The bank’s economist Pierfrancesco Mei compared the early estimates with decades of research on technology-related job losses, which found that the effects can extend beyond the initial layoff into lower pay and weaker earnings growth.
Goldman’s note, dated October 29, is not publicly available. Fortune reported its figures the following day, citing the note; the article’s source material says other coverage repeated Fortune’s numbers. The estimates should therefore be treated as reported findings, not independently verified results. No government survey directly asks workers whether AI caused their job loss.
To estimate AI-related displacement, Goldman reportedly examined more than 15,000 people laid off between 2016 and 2025 using Labor Department survey data. It combined occupational exposure to AI, the pace of AI adoption in an industry and whether a position was eliminated. Workers in the highest-scoring 30% were classified as “likely AI-displaced.” Goldman reportedly found no significant difference from other laid-off workers before 2023, when AI adoption began to rise.
For job losses since 2025, the reported analysis found that likely AI-displaced workers took about one month longer to find another job and were roughly 10 percentage points more likely to still be unemployed a year later than other displaced workers. Those aged 35 and older took about five weeks longer to return to full-time work, compared with about three weeks for workers under 35. These are comparisons between groups; they do not establish that AI alone caused each person’s job loss or subsequent unemployment.
Infographic 01 · What happens after the pink slip
The first AI layoffs look like 40 years of tech layoffs
A longer search, a smaller paycheck and slower growth for years. Goldman’s new estimate for AI-exposed workers lines up with its long-run history.
The first AI wave, since 2025
Likely AI-displaced vs. other laid-off workers, US. Note not public; figures as reported by Fortune.
Forty years of tech shocks
Workers displaced from technology-disrupted occupations, US.
“Comparable in magnitude” to earlier waves of technological change, whose damage was “long-lasting.”Goldman Sachs economist Pierfrancesco Mei, as quoted by Fortune
How Goldman spots “AI-displaced” workers: no survey asks about AI, so it scores each job loss by the occupation’s AI exposure, the industry’s AI adoption and whether the position was eliminated, and treats the top 30% as likely AI-displaced (more than 15,000 workers, 2016–2025). An estimate, not a count.
Sources: Fortune, 30 Oct 2026 (reporting the Goldman note of 29 Oct 2026) · Goldman Sachs, Top of Mind #149, 25 Jun 2026, p. 15 · Goldman Sachs Insights, 11 Sep 2026. pp = percentage points.
Powered by Thorsten Meyer AI
The Costs After an AI Layoff
The findings shift attention from the number of positions cut to the employment and earnings path after displacement. A longer job search can mean a longer spell without wages, while accepting work outside a former occupation may make it harder to regain previous pay or career progression. The note’s results are early estimates, but the historical comparison suggests those effects may outlast the initial period of unemployment.
Goldman’s public historical research found that workers displaced by earlier technologies took about a month longer to find work and lost more than 3% in real pay after reemployment, compared with negligible losses among workers laid off from stable or growing occupations. Over the following decade, their real earnings growth was nearly 10 percentage points lower. These longer-term findings concern past technology-related displacement, not a decade of observed outcomes for workers displaced by today’s AI systems.
The issue also matters for employers and policymakers weighing how to manage automation. If workers cannot readily move into comparable jobs, the costs may include prolonged unemployment and lower earnings, not just temporary disruption. Goldman’s historical data links retraining within three years to better outcomes, but that association does not by itself prove that training caused the improvement.
Popular right now
Vitamix 5200 Professional-Grade Blender, 64 oz. Container, Black
Bose QuietComfort Wireless Noise Cancelling Headphones, Sandstone
Breville Bambino Espresso Machine, Brushed Stainless Steel (BES450BSS)
As an affiliate, we earn on qualifying purchases.
Four Decades of Technology Displacement
Goldman’s comparison draws on research published in its June 2026 “Top of Mind” report and a September summary. Mei and colleagues Jessica Rindels and David Mericle examined workers across more than 300 occupations, using National Longitudinal Surveys that have followed over 20,000 people through their careers. They identified occupations more exposed to technology in each decade and tracked outcomes for people who held those jobs.
That research describes a recurring pattern: workers losing jobs to technology faced a slower return to work, a real-pay loss on reemployment and weaker earnings growth over time. It also found that displaced workers who retrained within three years had about two percentage points more real wage growth over ten years and a roughly 10-point lower chance of becoming unemployed again than those who did not retrain. The source material does not provide enough detail to treat those differences as proof that retraining alone produced the outcomes.
Government data offers a separate view of job loss, though it does not identify AI as the cause. Among long-tenured workers aged 25 to 54 who were displaced, 72.9% were working again by January 2026, compared with 57.3% of those aged 55 to 64. The source also cites Bureau of Labor Statistics figures showing that about 49% of long-tenured workers who returned to full-time work earned as much as or more than before, down from about 62% two years earlier.
““Comparable in magnitude””
— Pierfrancesco Mei, Goldman Sachs economist, as quoted in the reported note
Limits of the AI Job Estimate
The Goldman note is not public, and the AI-specific numbers in this account come from Fortune’s reporting, rather than direct review of the document. The method also identifies workers whose jobs are more exposed to AI and whose positions were eliminated; it cannot confirm that AI caused each layoff. The label “likely AI-displaced” reflects an estimate, not a verified reason for job loss.
The reported results cover a short period beginning in 2025. It remains unclear whether the longer job searches will persist, how outcomes vary by education, region or occupation, and whether those who find work will face lasting pay cuts. The source says Goldman expects AI-related job losses to accelerate if economic growth slows, but the scale and timing of any such increase are not established here.
Evidence to Watch in Coming Months
The next test will be whether future labor-market data shows the reported gap continuing as more workers are affected and more follow-up information becomes available. Goldman’s public research can provide a historical comparison, but its AI-specific estimates will need further observation and a clearer account of the underlying method to establish how robust they are.
Readers should also watch for updated government displacement figures and any public release of Goldman’s client note or additional analysis. Until then, the current evidence supports a cautious conclusion: early AI-related displacement may resemble earlier technology shocks in the difficulty of returning to work, while the size and duration of the AI-specific effects remain uncertain.
Key Questions
What did Goldman Sachs find about likely AI-related layoffs?
Fortune reported that workers Goldman classified as likely AI-displaced since 2025 took about a month longer to find work and were roughly 10 percentage points more likely to remain unemployed a year after displacement than other laid-off workers.
Are the Goldman figures independently confirmed?
No. The client note is not publicly available; the figures are reported through Fortune’s account. Goldman’s category is an estimate based on AI exposure, industry adoption and job elimination, not a direct confirmation that AI caused an individual layoff.
What does Goldman’s historical research say?
Its research on earlier technology-related job losses found that displaced workers took about a month longer to find work, lost more than 3% in real pay after reemployment and had nearly 10 percentage points less real earnings growth over the following decade than workers who were never laid off.
Did older workers have a harder time returning to work?
In the new figures reported by Fortune, workers aged 35 and older took about five weeks longer to return to full-time work, compared with about three weeks for those under 35. Separate government data cited in the source also shows lower reemployment rates among displaced workers aged 55 to 64 than among those aged 25 to 54.
What remains unknown about AI-related job losses?
The duration of the job-search and pay effects, differences by education or region, and the number of layoffs directly caused by AI remain unclear. The current AI-specific findings cover a short period and rely on an indirect classification method.
Source: ThorstenMeyerAI.com
