Workers displaced from jobs highly exposed to AI took about a month longer to find work than other displaced workers, according to Goldman Sachs research reported by Fortune. The figures suggest the effects of technological change may be showing up in jobseekers’ recovery, with older workers taking longer to get back to full-time work.
Watch Desk analysis
What happened
Goldman economist Pierfrancesco Mei analysed more than 15,000 people who lost jobs between 2016 and 2025, using newly released US survey data for 2023 to 2025. Fortune reports that workers Goldman classified as likely displaced from AI-exposed jobs took about a month longer to find a new job than other displaced workers, and were roughly 10 percentage points more likely to remain unemployed a year later.
The impact varied by age. Among those aged 35 and over, AI-exposed workers took about five weeks longer than other displaced workers to return to full-time work; for those under 35, the difference was about three weeks. Goldman’s classification is an estimate based on occupational AI exposure, industry adoption and whether the job was eliminated. The survey did not ask whether AI caused anyone’s job loss.
Why it matters
The numbers offer an early look at what happens after a job disappears, not just how many roles employers say they are cutting. A longer search and a greater chance of still being unemployed a year later are meaningful costs for workers, even if the causes of any individual layoff remain hard to pin down.
There are important limits. Goldman says its data likely captures the first wave of AI-displaced workers, and cautions that it is too early to assess effects across factors such as education and geography. Its analysis found no significant gap between the groups for job losses before 2023. Other research cited by Fortune has found no widespread, economy-wide job displacement associated with AI so far.
Our read
This is a useful warning light, not a verdict on AI’s effect on employment. The findings point to a potentially harder landing for some displaced workers, particularly older ones, but the label “AI-displaced” is a research classification rather than a confirmed reason for each job loss. The distinction matters: a careful early signal is worth discussing, while a neat causal story would be premature.
What to watch
- Whether later survey waves show the job-search gap widening, narrowing or persisting.
- How the results vary by age, industry, education and time spent in a job before displacement.
- Whether future research can distinguish AI-related job losses from other changes happening at the same time. Activity teaser: Goldman Sachs research suggests workers displaced from jobs it classifies as highly exposed to AI took about a month longer to find work than other displaced workers. Older workers took longer to return to full-time employment than younger ones. The findings are an early signal, not proof that AI caused each job loss: the survey did not ask people why they were laid off, and Goldman built its classification from exposure and industry data. The full story examines what the figures say, and what they do not.
Discussion spark: Should early research like this influence how governments support displaced workers, or is the evidence too uncertain to justify AI-specific policy?
Sources and evidence
- Source update (30 October 2026, 13:09 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.