Goldman Sachs says AI could push up to 9% of workers out of their jobs

Goldman Sachs has raised its estimate of how many workers AI could push out of their current roles. Its economists now say up to 9% could be displaced under their baseline forecast. That is about 15 million people whose jobs get disrupted. The bank says the change will likely come gradually, but a downturn could speed it up.

What changed

Last year, Goldman chief economist Jan Hatzius said AI could displace 6% to 7% of all US workers. The bank began floating a figure as high as 9% this summer. The new forecast assumes productivity, meaning how much work gets done per hour, rises 15% in the US and other developed economies.

Slower hiring is also a risk, since companies may hire fewer people as they adopt AI. But the report says the rise in unemployment will likely come more from outright job losses than from reduced hiring.

Where things stand now

The US is still early in this shift. Goldman estimates that less than a quarter of firms have built AI into their regular work. Hatzius wrote that AI labor disruption will likely "only gradually increase rather than delivering an abrupt labor market shock."

Layoff data shows AI is already a factor. Announced layoffs ticked lower in September, but AI was still the top cited reason, at 21% of all announced cuts. Tech had the most job losses of any industry, at 29% of cuts. Those figures come from Challenger, Gray & Christmas.

Why a downturn matters

The more worrying finding is about a weak economy. Goldman looked at past waves of automation during recessions. Hatzius wrote that a cyclical slowdown, meaning a normal dip in the economy, "would likely accelerate AI-related automation and job losses." In plain terms, companies under pressure to cut costs may move to AI faster, and that could make a downturn worse.

Who it affects

The forecast covers workers broadly, not one industry. Some researchers still fear AI could one day wipe out entry-level white-collar work, though the report notes that panic has eased slightly. Business owners may feel the change in how roles are structured and how many people they need.

What is not known yet

The material does not say which jobs are most at risk. It also does not say how many displaced workers would find new roles, or how long that would take. The 9% figure is a forecast with conditions attached, not a prediction of what will happen.