If you have been watching the monthly jobs numbers and wondering why they keep disappointing, Goldman Sachs may have part of your answer. June payrolls came in at just 57,000, less than half of what economists expected. April and May were revised down by a combined 74,000. And a growing number of economists are pointing to AI as a factor that is quietly reshaping who gets hired, who gets replaced, and who stops looking altogether.Goldman Sachs now says AI could displace roughly 15 million American workers over the next decade, or about 9% of the U.S. workforce. That estimate comes from Joseph Briggs, who co-leads Goldman Sachs Research’s global economics team. The bank’s revised methodology tracks not just people who are already unemployed, but everyone being steadily pushed out as employers automate more tasks each quarter.Goldman Sachs economist says AI job losses already show up in monthly dataBriggs laid out his case on the bank’s Exchanges podcast. He said AI tools are already pulling between 10,000 and 15,000 jobs out of monthly payroll growth, concentrated right now in tech, management consulting, and graphic design. Most employers have barely started deploying AI at any real scale.”9% of workers being displaced by AI would correspond to 15 million workers leaving or being displaced from their positions today and having to find new jobs,” he said.More Goldman Sachs:Goldman Sachs doubles down on stock market outlook for 2026Goldman hints at Fed’s next interest-rate bet under WarshGoldman Sachs has blunt message for AI stock investorsThe workers feeling it most right now are younger. Gen Z unemployment hit 8.3% as of June 2026, double the 4.2% national rate, according to workforce data. The New York Fed put the jobless rate for college graduates aged 22 to 27 at 5.6%. It is not that employers are running layoffs. They simply stopped backfilling. A June 2026 GMAC survey found that one in three employers has replaced entry-level roles with AI rather than hiring, Fortune reported.”I’m sure that we all know people who have had trouble finding jobs or a harder time than they would have normally following recent graduations,” Briggs said.The unemployment rate dipped to 4.2% in June, but that was mostly because 507,000 people stopped looking for work entirely, CNBC noted. This mechanically pulls the rate down without anyone actually finding a job.Goldman Sachs AI displacement forecast explained: how the bank got to 15 millionBriggs built his estimate on a historical pattern: Every 1% technology-driven productivity gain has tended to raise job destruction rates by about half a percentage point over the following two years. Goldman is forecasting that AI will deliver a 15% productivity boost at full adoption. That is where the 15 million figure comes from.Briggs is not saying this happens overnight. Spread over 10 years, he expects the unemployment rate rise to stay under one percentage point in any given year. But he has been warning since March that front-loaded job losses change that math fast. “The big story in 2026 in labor will be AI,” Briggs said at the time. “If we see some job losses pulled forward, that sets the stage for potential underperformance relative to our forecast, and that may lead the Federal Reserve to cut rates.” June’s 57,000-job reading fits that scenario.There is also a counterargument Briggs himself makes. A 5% pickup in the pace of U.S. job creation would be enough to reabsorb all AI-displaced workers. The economy already generates about 30 million jobs a year while destroying 29 million. The buffer exists; it just has to hold.
Goldman Sachs now says AI could displace roughly 15 million American workers over the next decade.Nagle/Getty Images
Which American workers face the biggest AI job displacement risk right nowCustomer service, back-office administration, and jobs built around doing the same cognitive task on repeat are where Goldman sees the most immediate exposure. A lot of workers in those roles assumed a desk job meant they were safe. That assumption is being tested.Briggs also made a point about who takes the hit. Lower-wage workers in predictable roles have nowhere easy to land when a job disappears. Higher earners usually have more options. He said income inequality is likely to widen through the transition, even if the overall jobs picture eventually stabilizes.Two MIT economists on the same podcast pushed back, though in different directions. Neil Thompson thinks real-world AI deployment will be slower than the models suggest. In regulated industries especially, companies still need data access and government clearances they do not have. His expectation is that most jobs get partially automated, not eliminated.Daron Acemoglu was less optimistic. He sees genuine net job losses in routine roles over the next five years, with more potential damage beyond that if AI investment continues to encourage replacing workers rather than making them more productive.What Goldman’s AI warning means for investors and S&P 500 companiesCorporate spending on AI has climbed through 2026. The efficiency case is real, and companies know it. But Goldman’s own data show that companies using AI to make workers more productive tend to outperform those using it purely to cut headcount.For stock market investors, the AI story has a side that rarely gets priced in. If enough workers lose jobs or see wages stagnate, consumer spending eventually slows. Companies that are too slow on AI adoption face competitors with lower costs. And if job losses accelerate faster than Goldman’s base case, the Federal Reserve could move on rates at a moment when inflation is already elevated.As the June jobs report made plain, the labor market is already under pressure, NPR noted. Whether AI is the main driver or one factor among several, 15 million workers finding themselves in different jobs over the next decade is a number worth taking seriously.Related: Nvidia CEO sends serious wake up call to all Americans