Eisman says S&P 500’s AI exposure tops 65%

Steve Eisman put artificial-intelligence exposure above 65% of the S&P 500 in an Oct. 9 podcast, according to a Benzinga trade-press article published Oct. 12.

Benzinga’s Oct. 12 article said Eisman’s warning lands after an approximately 14% gain for the S&P 500 in 2026, while Reuters reported more than 40% of index members have fallen.

Eisman, known for “The Big Short,” said broad index ownership now carries a heavier dependence on a single investment theme than buyers may recognize. “It’s mostly all one trade. And to be perfectly frank, that makes me nervous,” he said on the podcast published Oct.

9.

The caution is not a blanket rejection of AI-linked equities. Benzinga reported that Eisman owns Nvidia Corp. and GE Vernova Inc., both tied to the capital spending cycle around AI data centers and power infrastructure.

“I own Nvidia, for example, and I watch and I worry,” Eisman said in an Oct. 8 interview with In the Money’s Amber Kanwar.

Eisman’s distinction is between companies receiving AI infrastructure dollars and companies making the financial commitments behind the buildout. “I would be looking at the companies that are getting that cash flow,” he said in the Oct. 8 interview.

Benzinga reported that Nvidia’s revenue increased 106% from a year earlier to $96.2 billion in the quarter ended July 26. The article described Nvidia as a supplier of chips used in data centers, placing the company close to the spending flowing into AI computing capacity.

GE Vernova is exposed through the electricity and equipment side of the same buildout. Benzinga reported that the company’s companywide backlog was $176 billion at the end of June, and that data-center orders in its Electrification business were more than $5 billion in the first half of 2026, more than double its full-year 2025 level.

The concentration argument turns on what counts as AI exposure. Benzinga said Eisman estimated information technology at roughly 40% of the S&P 500. When Alphabet, Amazon and other technology-related companies are added, the share rises above 50%, according to Eisman’s count.

His estimate moves above 65% after including utilities and industrial companies that supply the infrastructure behind AI deployment. The SPDR S&P 500 ETF Trust tracks the index, so Eisman’s comments apply to investors using SPY as a broad U.S. equity allocation rather than as a direct technology bet.

Eisman also pointed to credit markets. Benzinga reported that he estimates more than half of this year’s corporate bond issuance is AI-related, which would tie lenders to the same capital spending cycle affecting equity investors.

The specific financing concern centers on large commitments by OpenAI and Anthropic. Eisman said those companies could struggle to finance their obligations, according to Benzinga, which could slow spending across suppliers if funding becomes more constrained.

There are limits to the warning. Benzinga reported that Eisman is not forecasting a collapse and remains long Nvidia and GE Vernova. The article also said he has partly hedged four unnamed AI-related positions while remaining mostly long.

Prediction-market pricing, as cited by Benzinga, does not show a high-probability crash scenario. Polymarket traders assigned about 6% odds to a qualifying AI industry downturn by year-end, under a contract requiring at least three specified stress events within a 90-day period rather than a standard market pullback.

For investors trying to avoid the theme entirely, Eisman pointed away from equities. “I would just buy short-term treasuries,” he said.

Source: trade-press article, Benzinga, Oct. 12, 2026