Goldman Sachs warns AI investment boom won’t last forever

Goldman Sachs is throwing cold water on one of the market’s most crowded trades. The bank’s chief economist, Jan Hatzius, is cautioning that the artificial intelligence investment frenzy is a multi-decade structural shift, not an infinite growth engine, and that its near-term economic impact has been dramatically overstated.

The core argument is disarmingly simple: most of the essential AI hardware is manufactured in Taiwan and South Korea, which means the massive capital expenditure flowing into AI infrastructure isn’t juicing US GDP the way investors seem to assume. In February 2026, Hatzius put it bluntly, stating that AI investments contributed “basically zero” to US GDP growth in 2025.

The trillion-dollar question

Goldman Sachs projected that global capital expenditure on AI would reach approximately $1.019 trillion in 2026. The US share of that total sits at around $581 billion.

Jim Covello, Goldman’s head of global equity research, has been even more direct than Hatzius. He remarked that AI’s economic prospects now seem “more questionable” than they did two years prior, citing unclear returns on the trillions in planned infrastructure spending.

Goldman’s August 2026 analysis estimated that AI capital expenditures would result in about $50 billion in incremental crowding-out effects on other business spending for the year.

A multi-decade shift, not a sugar rush

During a Goldman Sachs discussion on August 13, 2026, Hatzius framed the AI transition as something that will play out over decades, not quarters. He called for a balanced view that separates near-term hype from the technology’s genuine long-term potential.

On the labor front, Hatzius predicted stable unemployment rates despite AI’s potential to displace certain job roles.

The bank’s analysis also noted limited evidence so far that AI investments are crowding out other business expenditures at a dangerous scale. But it flagged two risks worth monitoring: high valuations across AI-adjacent stocks and the concentration of spending among a handful of dominant tech players.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.