By Thorsten Meyer | November 1, 2026 | ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Most stories about AI and jobs end on the day of the layoff. The headline counts the cuts, the company explains its reasons, and the story is over. For the people who lost the job, that day is the beginning. What happens next decides how much the layoff really costs: how long the search takes, what the next job pays, and whether it lasts.
This week, Goldman Sachs looked at exactly that. In a note to clients, its economist Pierfrancesco Mei compared the first workers who likely lost jobs to AI with four decades of Americans who lost jobs to earlier waves of technology. His conclusion, as reported by Fortune: the new group is starting to look a lot like the old one. And the old one carried the damage for years. Here is what Goldman found, what its public research and the government’s own data show, who is hit hardest, and what it means for companies and for the people on the receiving end.
Key takeaways
- The first AI-displaced workers search longer. Workers who likely lost AI-exposed jobs since 2025 took about a month longer to find new work than other laid-off workers, and were roughly 10 percentage points more likely to still be out of work a year later. That is a new Goldman note from 29 October, reported by Fortune; the note itself is not public.
- That matches 40 years of history. Goldman’s public research finds that workers who lost jobs to technology took about a month longer to find work, lost more than 3% in real pay when they did, and saw their real earnings grow nearly 10 percentage points less over the next decade than workers who were never laid off.
- Older workers are hit harder. In the new note, workers 35 and older took about five weeks longer than comparable laid-off workers to get back to full-time work; under 35, about three weeks. Government data for all laid-off workers shows the same gap: 72.9% of long-tenured workers aged 25 to 54 were working again by January 2026, but only 57.3% of those 55 to 64.
- The pay hit is growing. Of long-tenured workers who found full-time work again, about 49% earned as much as before or more, down from about 62% two years earlier (Bureau of Labor Statistics).
- Retraining is linked to better outcomes. In Goldman’s historical data, technology-displaced workers who retrained within three years had about 2 percentage points more real wage growth over ten years and a roughly 10-point lower chance of becoming unemployed again.
- A recession would make it worse. Goldman finds displacement costs are larger in downturns, and its economists expect AI-related job losses to speed up when the economy slows.
Products Worth Considering
Apple Watch SE 3 GPS 40mm Smartwatch
Bose QuietComfort Wireless Noise-Cancelling Headphones, Petal Pink
DJI Osmo Mobile 8 Gimbal with Mic Mini Transmitter, Infinity Black
As an affiliate, we earn on qualifying purchases.
What Goldman found this week
The new note was published on Thursday, 29 October 2026. Goldman sends notes like this to clients, so I could not read it myself. The details below come from Fortune’s report the next day, by business editor Nick Lichtenberg, which quotes the note directly. Other outlets that covered it repeat Fortune’s numbers. Treat the figures in this section as reported, not independently confirmed.
First, a problem: no government survey asks people whether AI cost them their job. So Goldman built an estimate. It took the Labor Department’s survey of laid-off workers, the Displaced Worker Supplement to the Current Population Survey, and looked at more than 15,000 people who lost a job between 2016 and 2025. For each job loss, it combined three things into a score: how exposed the occupation is to AI, how fast the industry is adopting AI, and whether the position itself was eliminated. The top 30% count as “likely AI-displaced”. As a check, Goldman found no significant gap between this group and other laid-off workers for job losses before 2023, the year AI adoption began to rise.
What Fortune reports Goldman found for job losses since 2025:
- A longer search. About one month longer to find a new job than other displaced workers.
- More people stuck. Roughly 10 percentage points more likely to still be unemployed a year after losing the job.
- The age gap. Workers 35 and older took about five weeks longer to get back to full-time work; those under 35 about three weeks.
- A rougher path for older workers. Younger AI-displaced workers cycled through fewer temporary jobs on the way back and were less likely to need a second job next to the new full-time one.
- A risky landing. AI-displaced workers were more likely to leave their old occupation and industry altogether. But when they were rehired, more of them, “especially older ones,” moved into roles with similar or even higher AI exposure.
Mei called these effects “comparable in magnitude” to earlier waves of technological change, and called the damage from those waves “long-lasting.” He also wrote that it is “still too early to assess the impact along all of these dimensions,” such as education and region.
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 40-year pattern: what Goldman’s public research shows
The comparison rests on research Goldman has published openly. In its June 2026 “Top of Mind” report and a September summary, Mei and his colleagues Jessica Rindels and David Mericle tracked people in more than 300 occupations since the 1980s. In each decade they identified the jobs most exposed to technology, then followed the individuals who held them through the National Longitudinal Surveys, which have followed over 20,000 workers through their careers.
Their findings, in plain words:
- The way back is slower. Workers who lost jobs to technology took about one month longer to find new work than workers laid off from stable or growing occupations.
- The next job pays less. On reemployment, they lost more than 3% in real pay, against “negligible” losses for workers from stable occupations.
- The gap keeps growing. Over the following decade, their real earnings grew nearly 10 percentage points less than those of workers who were never laid off, and about 5 points less than those of other laid-off workers.
- The next layoff comes sooner. Their chance of another spell of unemployment stayed higher, especially in the first five years.
- It reaches beyond work. People displaced early in their careers, between 25 and 35, built wealth more slowly, largely because they bought homes later, and were slightly less likely to form a household.
Why? Goldman’s answer is “occupational downgrading.” The same technology that eliminated a job also made the worker’s skills less valuable. So many moved into work with more repetitive, rule-based tasks and fewer analytical or people-facing demands. Think of an experienced typesetter in the 1980s: the skill was real, but the market for it had just shrunk.
Goldman draws a clear line from this to AI: “Our analysis suggests that, similarly to previous waves of technological change, AI-driven displacement could impose lasting costs on affected workers, worsening labor market outcomes for several years.”
Why economists call it a “scar”
Goldman is not the first to find this. Economists have measured the long shadow of a layoff for more than thirty years. In a classic 1993 study, Louis Jacobson, Robert LaLonde and Daniel Sullivan found that long-tenured workers who left distressed firms suffered long-term earnings losses averaging 25% per year. Steven Davis and Till von Wachter, using Social Security records from 1974 to 2008, found that men displaced in mass layoffs lost on average 1.4 years of their former earnings when unemployment was below 6%, and 2.8 years when it was above 8%.
That last number matters for AI. Goldman’s own research finds the same timing effect: workers displaced by technology during downturns spend about three more weeks unemployed and face a roughly 5 percentage point higher risk of later unemployment or leaving the workforce. In a separate note on 22 October, reported by Fortune, Goldman’s Joseph Briggs wrote that automation-related labor disruption “is limited when the overall economy is healthy but accelerates during cyclical slowdowns.”
Who is hit hardest
Age is the clearest dividing line, and the government’s numbers show it for all laid-off workers, not only the AI-exposed. The Bureau of Labor Statistics published its latest Displaced Worker survey on 27 August 2026:
- 7.4 million workers were displaced from 2023 to 2025, up from 6.3 million in the previous survey covering 2021 to 2023.
- Of the 3.3 million who had held their job for at least three years, 66.1% were working again in January 2026 and 18.3% were unemployed (16.1% two years earlier, a change the BLS calls little different).
- By age: 72.9% of those aged 25 to 54 were reemployed, 57.3% of those aged 55 to 64, and 38.6% of those 65 and older.
- By pay: of those back in full-time jobs, about 49% earned as much as before or more, down from about 62% in the previous survey.
- By reason: 44.4% lost their job because their position or shift was abolished, 32.6% because their plant or company closed or moved, and 22.9% because there was not enough work.
Infographic 02 · Age decides the road back
The older you are, the longer the way back
The Labor Department’s survey of laid-off workers: who works again, and at what pay.
Back at work by January 2026
Long-tenured workers (3+ years in the job) displaced 2023–2025, share reemployed, by age
- Ages 25–5472.9%
- Ages 55–6457.3%
- Age 65 and older38.6%
The pay hit is growing
Reemployed full-time who earn as much as before or more
- Survey of January 2024displaced 2021–2023≈ 62%
- Survey of January 2026displaced 2023–2025≈ 49%
7.4 million
workers displaced 2023–2025, up from 6.3 million in 2021–2023
18.3% unemployed
of long-tenured displaced workers in January 2026 (16.1% two years earlier)
44.4% job abolished
position or shift abolished; 32.6% plant closed or moved; 22.9% not enough work
These are government figures for all displaced workers, not only AI-exposed ones. Goldman reports the same age pattern for its AI-displaced group: about five weeks longer back to full-time work at 35+, about three weeks under 35 (as reported by Fortune).
Source: BLS, Worker Displacement 2023-2025, 27 Aug 2026 (Displaced Worker Supplement, January 2026). Shares as published.
Powered by Thorsten Meyer AI
Age is not the only factor. In Goldman’s historical data, younger, college-educated and urban workers who lost jobs to technology had roughly half the cumulative earnings losses of other technology-displaced workers, because they could move more easily into less routine roles. Workers with shorter tenure also did better, likely because their skills were less tied to one occupation. The new note sees the same split so far: according to Fortune, Mei wrote that younger AI-displaced workers “may be able to adapt more flexibly to AI disruption” and called that “an encouraging sign given concerns that AI-driven job losses could disproportionately affect recent graduates.”
There is a catch. Goldman’s data covers only people who had a job to lose. Young people who never get hired in the first place do not show up as displaced. That front-door problem is a separate story, which I looked at in Computer Science Enrollment Fell 12.4% This Fall. Is AI Really Taking the First Job?.
Is AI really the reason?
Two earlier pieces on this site set the context. In Is AI Really Behind the Layoffs? I showed that US employers named AI in 120,136 announced job cuts from January to September, while surveys find AI behind far fewer actual layoffs. In Is Automation Now an Imperative? I looked at why governments and companies keep shrinking, with AI more often the tool than the reason.
Goldman’s note sidesteps that debate in an interesting way. It does not rely on what companies say. It estimates who probably lost work to AI from the type of job, the industry and whether the position disappeared. That makes it a statistical estimate, not a count. Some people in the “AI-displaced” group surely lost their jobs for other reasons. But it also means the result does not depend on a press release. Whatever companies call it, the people in AI-exposed jobs who were laid off since 2025 are having a harder time getting back.
How big could this get? In June, Goldman’s Joseph Briggs estimated that the AI transition could displace over 9% of the US labor force, around 15 million workers, spread over a ten-year adoption period. He expects most of them to find new work within a year and new jobs to absorb the losses over time. The scar research is about what that year, and the decade after it, costs the people involved.
What this means for companies
- The layoff is not the end of your cost. If the people you let go struggle for years, that shows up in your reputation, in the remaining team’s trust and in how hard it is to hire the next time. Plan the exit as carefully as the cut.
- Retraining before the cut beats severance after it. Goldman’s data says retraining within three years of a job loss improves pay and stability over the next decade. The cheapest moment to retrain someone is while they still work for you. Internal moves into roles that use AI, rather than compete with it, keep skills and knowledge in the company.
- Older employees need a different plan. They take longer to find work, more often land in lower-paid roles, and, according to Goldman, often move into jobs just as exposed to AI as the old one. Career advice, longer notice and help with the search matter most for them.
- Say why, consistently. If AI is the reason, say so and show the plan for the people affected. If it is not, do not use it as a label. Inconsistent reasons are now easy to compare, as the AI-washing analysis showed.
- Prepare for the downturn now. If AI cuts cluster in the next slowdown, as Goldman’s economists expect, the people displaced then will face the steepest losses. Transition programs built in good times are easier to fund than in bad ones.
What this means for workers
This is general information, not career or financial advice. Your situation depends on your job, your field and where you live.
- Count the search in months, not weeks. If your role is highly exposed to AI, the historical data suggests planning for a longer search than usual. A larger cash buffer and knowing your severance and unemployment rules in advance buy you time to choose well instead of taking the first offer.
- Watch where you land. Goldman found that many rehired AI-displaced workers, especially older ones, moved into roles with similar or higher AI exposure. Before you accept a new job, ask how AI is changing that role. Otherwise you may face the same risk again in a few years.
- Retrain early, and for real. In Goldman’s data, retraining meant at least four weeks of training within three years of the job loss, and it was linked to better pay and fewer new layoffs. Retrained workers tended to move into roles with more analytical work that complements technology, rather than repeating routine steps.
- Your age is a factor, not a verdict. Older workers face a harder road on average. Goldman also found that workers whose skills were less tied to one occupation did better. Making your experience portable, for example by describing what you did in terms other industries understand, is one way to work on that.
- Do not wait for the layoff. The best time to learn the AI tools your field actually uses, and to build contacts outside your current employer, is while you still have a job.
Infographic 03 · What cushions the fall
Retraining helps. A recession hurts. Both are measurable.
What decades of data say about softening a technology layoff, and what companies and workers can do before it happens.
For companies
- Retrain before you cut. The cheapest moment to retrain is while people still work for you.
- Move people inside. Into roles that use AI rather than compete with it.
- A different plan for older staff. Longer notice, search help, career advice.
- One reason, everywhere. Press release, filing and staff memo should say the same.
- Build it in good times. Goldman expects AI job losses to speed up in a slowdown.
For workers
- Plan the search in months. A cash buffer buys time to choose well.
- Check the next job’s AI exposure. Many rehired workers landed in roles just as exposed.
- Retrain early. In Goldman’s data: at least four weeks within three years of the job loss.
- Make experience portable. Describe it in terms other industries understand.
- Start before the layoff. Learn the AI tools your field actually uses.
- Job lossThe skill loses market value
- Longer searchAbout a month more for tech-displaced workers
- DowngradeMore routine, lower-paid work
- Next riskHigher chance of another layoff, esp. first 5 years
Retraining results are associations in historical data, not a guarantee. The 1.4 and 2.8 years are present-value losses for men displaced in mass layoffs when US unemployment was below 6% or above 8% (Davis/von Wachter). Long-tenured workers leaving distressed firms lost about 25% of earnings per year in the long run (Jacobson/LaLonde/Sullivan). General information, not career or financial advice.
Sources: Goldman Sachs, Top of Mind #149, 25 Jun 2026, p. 15 · Goldman Sachs Insights, 11 Sep 2026 · Davis, von Wachter (NBER w17638) · Jacobson, LaLonde, Sullivan (1993).
Powered by Thorsten Meyer AI
What we don’t know yet
- The note itself. I could not read Goldman’s 29 October note. The AI-specific numbers come from Fortune’s report; other outlets repeat Fortune. The 40-year findings, by contrast, are in Goldman’s public research.
- Who exactly is “AI-displaced”. Goldman’s group is an estimate built from job type, industry and whether a position was cut. Fortune does not report a list of occupations or a breakdown by industry.
- The long run. The AI group has been watched for a year or two at most. Whether their pay falls behind for a decade, like earlier waves, is a forecast based on history, not yet a measurement.
- Everything outside the survey. People who never got hired, people who left the workforce early, and freelancers whose work dried up are not fully captured.
- The economy-wide picture. Other researchers still see no broad, economy-wide jump in unemployment linked to AI. Goldman’s finding is about the people who did lose jobs, not about how many did.
The bottom line
The question “will AI take jobs?” gets most of the attention. Goldman’s note points to a quieter question that may matter more for the people involved: what happens after? For forty years, the answer for workers displaced by technology has been a longer search, a smaller paycheck and a higher chance of the next layoff. The first AI-displaced workers appear to be on the same path, and older workers most of all. History also shows what helps: retraining that comes early, moves into work that complements the technology, and a cushion that buys time. Companies and workers can start on all three before the next recession makes it harder.
Sources
- Fortune, Nick Lichtenberg: “Goldman says AI layoffs are starting to leave a ‘long-lasting’ scar, just like the last 40 years of history”, 30 October 2026 (reporting on a Goldman Sachs note by Pierfrancesco Mei published 29 October 2026). fortune.com
- Goldman Sachs Research, Top of Mind Issue 149, “An AI Job Apocalypse?”, 25 June 2026, including Pierfrancesco Mei, “AI’s uneven toll on workers” (p. 15) and Joseph Briggs on displacement (pp. 9–10). goldmansachs.com (PDF)
- Goldman Sachs Insights, “How Automation Impacts the Lives and Careers of Displaced Workers”, 11 September 2026. goldmansachs.com
- US Bureau of Labor Statistics, “Worker Displacement: 2023-2025”, USDL-26-1423, 27 August 2026. bls.gov
- Louis S. Jacobson, Robert J. LaLonde, Daniel G. Sullivan, “Earnings Losses of Displaced Workers”, American Economic Review 83(4), 1993, pp. 685–709. RePEc (working paper)
- Steven J. Davis, Till M. von Wachter, “Recessions and the Costs of Job Loss”, NBER Working Paper 17638, 2011. nber.org
- Fortune, “AI could displace one in 11 workers, Goldman says, but the layoffs are likely waiting for a recession”, 22 October 2026. fortune.com
This article is for information only. It is not career, legal, financial or investment advice. Mentions of Goldman Sachs refer to its published research and reported notes, not to its shares.
