Steve Eisman built his reputation shorting subprime mortgages before the 2008 financial collapse. Now he is pointing at what he sees as a similar crack in the AI boom.
The problem, he says, is that the biggest tech companies in the world are placing one massive bet. They are betting that two AI startups keep winning.
Eisman, who hosts "The Real Eisman Playbook" podcast and formerly managed portfolios at Neuberger Berman, shared his concerns Tuesday on CNBC's "Fast Money." His math is straightforward.
Those two companies also make up 25% to 35% of the cloud revenue at those same businesses.
Cloud revenue is the money tech giants make from renting out their computing power and data centers to other companies. In plain terms, a huge slice of the AI economy is running through just two doors.
"The futures of these massive companies, in a sense, are a bet that OpenAI, Anthropic are going to succeed," Eisman said.
Get the free Always Be Buying eBook and learn the simple system for building wealth on any income
That concentration is what worries him. If one of those two stumbles, the ripple would hit the biggest names in the stock market.
Eisman sees one threat bigger than the rest, and it comes from China.
Chinese open-source and open-weight AI models are much cheaper to use, and they appear to be taking market share. Open-source means the underlying code is public, so developers can use it, tweak it, and build on it without paying a licensing fee.
Eisman described the biggest vulnerability as the possibility that something goes wrong for Anthropic and OpenAI while cheaper Chinese alternatives are already gaining ground. If those models start capturing significant market share, he warned, the result could be a major price war that creates serious problems for the industry.
A price war in AI would mean companies slashing what they charge for AI services. That sounds great for customers, but it would squeeze the profits of the tech giants who have poured enormous sums into building the infrastructure.
Eisman is not the only "Big Short" alum sounding alarms. Michael Burry, another investor featured in the book and film, is even more pessimistic. He doubts that much of AI demand comes from real end customers, arguing that much of it is funded via what he has called "circular arrangements," where companies are essentially paying each other in a loop. Burry has shorted several of the largest AI winners, such as Nvidia, and taken positions against the broader semiconductor sector.
This debate is not academic. It sits at the center of the biggest question in markets right now, which is whether the massive AI buildout will actually generate returns.
Tech giants are pouring massive amounts into data centers and chips. If the revenue never shows up, or if cheaper Chinese models undercut the whole business, the stocks that led the market higher could fall hard.
Eisman's warning adds a prominent voice to that argument. His reputation stems from that prescient subprime short, which lends weight to his words, though being right once does not guarantee being right again.
The timeline matters here. This is a story that will play out over years, not weeks, as more earnings reports show whether AI revenue is growing fast enough to justify the costs.
For everyday investors, the takeaway is not about panic or a rush to sell. It is about understanding that when a few companies carry that much weight, the whole market gets bumpier. The AI boom has real momentum, but it also has real concentration risk, and that is worth knowing whether you own tech stocks, index funds, or just a retirement account that quietly holds a piece of everything.
Download the free Always Be Buying eBook and start putting your money to work today