Goldman’s Separate AI Estimates Put Displacement at 6%–7% and 9%

Goldman Sachs published two distinct estimates of artificial intelligence’s potential effect on U.S. workers: one projected 9% would be reallocated to new positions over 10 years, while another estimated 6% to 7% could be displaced over the next decade.

The figures come from separate research teams and the analyses do not reconcile the difference. They do not establish a revised forecast or a single 6% to 9% range.

In a July 2 transcript, Goldman Sachs Research economist Joseph Briggs discussed a scenario in which full AI adoption eventually lifts productivity by 15%. He said the 9% figure would represent 15 million workers: “Nine percent of workers being displaced by AI would correspond to 15 million workers.” The figure counts workers expected to shift into new positions over a decade; it does not forecast 15 million people remaining jobless.

A Sept. 11 analysis by Goldman economists Pierfrancesco Mei, Jessica Rindels and David Mericle put potential displacement at 6% to 7% over the next decade. That analysis warned the effects could be “substantially larger” if displacement happens during a recession.

Briggs has also said technology-created jobs could eventually absorb workers displaced by automation, with the pace of job creation shaping the transition. The July estimate was covered in an earlier report on Goldman’s 9% projection.