While the latest wave of AI-linked layoffs has put job seekers—and even the Federal Reserve—on high alert, a new survey from Goldman Sachs suggests the real AI labor meltdown is still to come.
The report, which surveyed more than 100 Goldman Sachs investment bankers, found that only 11% of their clients across industries such as tech, industrials, and finance were actively cutting employees as a result of AI. Instead, 47% of the bankers reported their clients were disproportionately using AI to boost productivity and revenue, while only a fifth were mostly using the tech to cut costs.
“AI use has so far been more skewed toward raising productivity/revenue than reducing costs,” wrote analysts led by Goldman Sachs chief economist and head of global investment research Jan Hatzius.
The catch: A much higher percentage (31%) of tech, media, and communications companies were cutting jobs because of AI. This caveat is reflected in the spate of mass layoffs that large tech companies have conducted over the past couple of months.
Amazon earlier this week was the latest—laying off 14,000 middle managers as the company prepares for a new world of advanced AI with a “leaner” workforce. Other companies such as Salesforce and tech-focused consultancy Accenture have together added tens of thousands of workers to the pile of AI-related layoffs in the past few months. The headlines have been so bleak that Fed Chair Jerome Powell said the Federal Reserve is watching carefully.
While companies may not be laying off workers now, bankers believe more layoffs could occur in the next few years. Over the next year, the bankers predict their clients will push forward a 4% general decrease in headcount, while over the next three years, those headcount reductions could skyrocket to 11%.
The worst-affected category for future layoffs is financial institutions, which bankers predict could see a 14% reduction in general headcount over the next three years. Tech, which has been among the quickest to adopt AI, could see slightly lower cuts of 10%.
“The relatively fast increase in expected adoption and headcount reductions over the next three years highlights that AI impacts on the U.S. labor market could arrive sooner than expected,” wrote the Goldman analysts.