Wall Street research and advisory firm Gartner is standing by its forecast that $3.5 trillion in AI investment could be on the rocks due to profitability concerns among the top AI models.
Gartner's forecast, issued last week, helped send shockwaves through the AI trade. The firm's chief forecaster, John Lovelock, told CNBC Monday that returns are not yet backing up the full scale of investment.
Consolidation and takeovers among AI-dedicated companies known as neoclouds could ensue, due partly to a mismatch in timing, he added.
"At the moment, we have more compute capacity than we have demand, and we have a lot of the tokens and processing power being given away in freemium solutions," he told CNBC. "Everyone in the AI race is running for a last-man standing position. Pretty much every market that Gartner is tracking in our AI race is likely to collapse down to a duopoly."
That duopoly for the main commercial frontier models consists of Anthropic and OpenAI. The computing-power asset base that's being constructed by neoclouds to serve that duopoly requires a degree of cash flow for those businesses that isn't yet scaling in proportion to the investment, Lovelock said.
Neoclouds would not be going out of business en masse, Lovelock reiterated, but could be subject to consolidation, such as takeovers and mergers.
As a result of their profitability analysis, Gartner predicted last week that 80% of neoclouds, which are AI-dedicated computing power providers, could cease functioning as independent companies.
Neoclouds were hit hard last week, helping to pull the rest of the stock market down with them.
CoreWeave dropped 25%; Nebius dropped 25%; and IREN fell 16%. The S&P 500 finished down 2% and the Nasdaq Composite fell 4% for the week ending Oct. 16, while the Dow Jones Industrial Average finished up 0.7% as investors rotated away from technology stocks.
CoreWeave was up about 2% in midday trading on Monday, while Nebius was up 3.3% and IREN was ahead 5.6%.
Analysts for Wolfe said in a Sunday note that the Gartner report likely contributed to the market stress.
"By the end of [Wednesday], the market appeared to already be on edge when a report from Gartner dropped, warning that more than $3.5tn in AI infrastructure spending could be exposed to impairment by 2030, with neocloud providers particularly vulnerable," Stephanie Wolfe at Roth wrote to clients.
Lovelock also said Monday that AI is a technology as opposed to a market, comparing it to table salt that will be sprinkled into existing commercial operations rather than creating an inherently new field in which companies will sell differentiated and competing products.
In their Wednesday report, Gartner researchers found that overall "AI revenue will be insufficient to support profitable returns on all planned AI capital expenditures" between 2027 and 2030.
The finding came as revenue growth for frontier algorithm company Anthropic decelerated, falling from 140% growth in the second quarter to 45% growth in the third.
The quarterly revenue level of $16.8 billion for Anthropic, which many tech analysts view as one of the linchpins of the AI trade, is quite small relative to the more than $5 trillion that hyperscalers are expected to invest through 2030, the pace of which is increasing.
OpenAI annual revenue expectations missed by $20 billion the week before, Roth at Wolfe noted.
Gartner hasn't been alone in warning that investors' math for the AI buildout may be off. Jefferies' global head of equity strategy Chris Wood has been warning about capital destruction in AI since at least September. Famous investor Paul Tudor Jones told CNBC last week that macroeconomic considerations would lead to a challenging year for the stock market in 2027.
"[If] we sum up all the buybacks, M&A, all the positive things that have powered this market so much … say since 2009, if I sum up all those and then I subtract from that the IPOs, secondaries, unlocks, et cetera, then you've got certainly the most negative flow picture since 2009," he said.