Gartner Stands by Warning That AI Returns Don't Yet Justify Neocloud Build-Out

Gartner is sticking with its forecast that $3.5 trillion in AI investment could be at risk because the top AI models aren't yet profitable enough. The firm's chief forecaster, John Lovelock, told CNBC on Monday that returns aren't backing up the scale of spending.

He said there is currently more compute capacity than demand, with plenty of tokens and processing power given away in freemium offerings. "Everyone in the AI race is running for a last-man standing position," he said, predicting most markets Gartner tracks will collapse to a duopoly - which for frontier models means Anthropic and OpenAI.

Neoclouds building computing power for that duopoly need cash flow that isn't scaling with investment, Lovelock said. He stressed they wouldn't go out of business en masse but could face consolidation through takeovers and mergers. Gartner predicted last week that 80% of neoclouds could cease operating as independent companies.

The forecast shook the AI trade. CoreWeave and Nebius each fell 25% and IREN dropped 16% for the week ending Oct. 16, while the S&P 500 lost 2% and the Nasdaq Composite 4%. By midday Monday, CoreWeave was up about 2%, Nebius 3.3% and IREN 5.6%.

Roth's Stephanie Wolfe wrote that the Gartner report likely contributed to market stress, warning that more than $3.5 trillion in AI infrastructure spending could face impairment by 2030. Gartner researchers found AI revenue would be insufficient to support profitable returns on all planned AI capital expenditures between 2027 and 2030.

Anthropic's revenue growth slowed from 140% in the second quarter to 45% in the third, reaching $16.8 billion - small next to the more than $5 trillion hyperscalers are expected to invest through 2030. Lovelock compared AI to table salt, sprinkled into existing operations rather than creating a new field of competing products.