Goldman sees AI displacement risk rising during economic downturns

Goldman Sachs economist Joseph Briggs estimates that AI could force roughly 9% of US workers to change jobs over a decade. Separate research from the bank finds that recessions intensify the damage from technological displacement as businesses eliminate routine work.

In a July discussion, Briggs put the potential displacement at about 15 million workers. His calculation assumes widespread AI adoption eventually raises productivity by 15%. He described a gradual movement between jobs, rather than permanent unemployment for everyone affected.

The transition can still leave lasting costs. In research published in September, Goldman said workers displaced by technology historically took longer to find employment and suffered greater earnings losses than other displaced workers. Those consequences were worse when job losses occurred during a recession.

Briggs expects new occupations and expanding demand to absorb displaced workers over time. The unresolved issue is whether those opportunities develop quickly enough to offset the losses during the transition. Goldman’s research also identifies retraining as one way to improve subsequent earnings and employment stability.