'AI Labor Disruption': Goldman Sachs Forecasts Layoffs Will Displace Nine Percent of Staff Globally

Goldman Sachs warned on Thursday, 22 October 2026, that artificial intelligence could displace around 9% of US workers over a decade. Its latest analysis suggests an economic downturn could accelerate automation-related job losses.

The warning builds on an earlier forecast by Joseph Briggs, who leads the bank's global economics team. In July, Briggs estimated that about 15 million US workers could leave their existing positions as AI adoption spreads. That is a projection of job displacement, not permanent unemployment.

Goldman's baseline assumes AI eventually raises productivity by 15% in the US and other developed economies. The outcome remains uncertain, and the forecast does not mean 9% of employees worldwide will be made redundant.

Why Goldman Sachs Expects AI Job Disruption to Unfold Gradually

Goldman's economists expect the effects to spread as employers introduce technology capable of taking over certain tasks. The bank is not predicting that millions of workers will suddenly lose their jobs.

In a July discussion published by Goldman Sachs, Briggs described the expected transition as lasting 10 years. He said the effect on unemployment in any individual year would depend on how evenly displacement was spread. Workers may move into new positions, although that does not guarantee a smooth transition.

The bank previously estimated that widespread AI adoption could displace 6% to 7% of US workers. Its later projection of about 9% reflects a more substantial potential reshaping of employment, rather than an immediate count of redundancies.

Existing evidence suggests that disruption will not fall equally across the workforce. Goldman Sachs Research reported in September that hiring had weakened in some AI-exposed fields, including software publishing, advertising and call centres.

Its research also identified particular pressure on junior employees seeking work in affected industries. Nevertheless, the economists described the impact on economy-wide hiring as limited so far. That distinction matters for graduates facing tougher entry routes even while total employment remains comparatively resilient.

There may be longer-term costs for people who lose their jobs. Goldman's September study of past automation found that displaced workers often faced more difficult employment transitions and greater earnings losses than other people who lost work.

Retraining can help, according to that study, but it is no guarantee of another position. Someone moving into a different occupation may need time to build skills while managing the financial consequences of unemployment.

Why Goldman Sachs Warns a Recession Could Make AI Layoffs Worse

The bank's October analysis raises a more troubling possibility. Businesses facing weaker demand may accelerate automation as they look to reduce labour costs, potentially making a slowdown worse for workers.

That concern draws on earlier periods of technological change, when displacement tended to intensify during economic downturns. It is a warning about what could happen, not evidence that a fresh recession or a sudden wave of AI redundancies is inevitable.

The latest US job-cut figures show why timing and definitions matter. Challenger, Gray & Christmas recorded 43,281 announced job cuts in September, 18% fewer than in August.

AI was cited in 3,961 September job-cut announcements, about 9% of that month's total. Across January to September, however, employers attributed 120,136 planned cuts to AI, approximately 21% of all announced reductions.

Technology companies announced 10,799 cuts in September. Their 165,925 announced cuts across the first nine months of 2026 represented 29% of the year-to-date total, the largest share of any industry.

Those figures measure employers' announcements, not confirmed cases of workers being replaced by software. AI being cited as a reason does not establish that automation alone caused every planned reduction.

The picture for recruitment is also mixed. Challenger recorded 210,612 planned hires from January through September, 3% more than in the same period of 2025.

September told a different story. Employers announced 90,787 planned hires, the lowest September total since 2011, according to Challenger. The figure was 23% below September 2025.