Goldman Sachs urged clients to tread carefully in the red-hot artificial intelligence trade, cautioning that lofty expectations could collide with reality.
In a note late Thursday, the bank’s strategists said AI remains a powerful long-term theme but warned that near-term risks are mounting.
The concern: a potential cooling in capital spending by big technology firms after a year of aggressive investment in data centers and chips.
Goldman Sachs warnings amid the AI investment boom
Investor enthusiasm for artificial intelligence stocks may be nearing its peak, according to Goldman Sachs strategist Ryan Hammond, who sees the sector shifting toward a more demanding phase of scrutiny.
Hammond noted that the current AI investment cycle has been powered largely by “hyperscaler” heavyweights like Amazon, Microsoft, Alphabet, Meta and Oracle, which together poured roughly $368 billion into AI-related capex and research in 2025, a record surge.
But the strategist warned that this second phase of aggressive buildout may soon give way to a third stage, in which investors look beyond infrastructure spending and demand clearer evidence of earnings gains.
The shift underscores the risk that AI valuations, which have already soared on expectations, could face pressure unless companies begin to show near-term financial payoffs.
Hammond explained:
While we expect the AI trade will eventually transition to Phase 3, investors will likely require evidence of a tangible impact on near-term earnings to embrace these stocks. Unlike Phase 2, there will likely be winners and losers within Phase 3.
Investor exuberance over artificial intelligence is showing cracks as Wall Street questions whether corporate earnings can keep pace with the hype.
Nvidia stock tumbled almost 6% after investors took a harder look at its guidance, while Salesforce and newly listed Figma also slid when results failed to justify rich valuations.
The pullbacks have fueled broader doubts about whether the sector’s meteoric rise can be sustained without clearer signs of near-term profit growth.
Risk of a capital expenditure slowdown
Goldman Sachs is cautioning that any pullback in artificial intelligence spending by Big Tech could hit markets hard.
If the “hyperscaler” companies were to scale back capital investment to 2022 levels, a scenario Goldman labeled “extreme,” the impact could be severe.
The firm estimates such a retreat would wipe out roughly 30% of the $1 trillion in S&P 500 sales growth projected for 2026.
That, in turn, could compress valuation multiples by 15% to 20%, dealing a heavy blow to AI-linked shares and broader equity benchmarks.
Goldman Sachs noted that while valuations for leading tech companies remain elevated, they are still well below the extremes of past market bubbles.
The firm highlighted that the five largest US technology stocks currently trade at roughly 28 times earnings, compared with a 40x multiple at the height of 2021’s rally and about 50x during the dot-com era in 2000.
Goldman credited this relative restraint to the substantial real earnings now being generated, largely supported by hyperscaler capital investment in AI infrastructure.
The bank suggested that unlike previous speculative cycles, today’s valuations are underpinned by stronger fundamentals, though risks remain if spending growth slows.
The post Goldman Sachs sounds alarm on AI bubble risk: why analysts are worried appeared first on Invezz
