Nvidia Rounds Up Six Financial Giants for $500 Billion AI Financing Push as Bank CEOs Flag Lending Risk

The AI buildout has a funding problem, and Wall Street is being asked to solve it.

In August, Nvidia brought together six financial heavyweights (Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR) to set up financing platforms that could mobilize more than $500 billion for companies building AI infrastructure, according to an analysis published by Fortune on Oct. 18. The company that sells the chips is now helping arrange the money to buy them.

The spending is outrunning cash flow

The numbers are enormous. The five largest U.S. hyperscalers are on track to spend $730 billion in capital expenditures in 2026, Fortune reports. They are expected to finance up to a third of that with debt.

That is a change for companies that spent half a century selling software with almost no marginal cost. AI runs on factories, advanced chips, substations and huge amounts of electricity. Even the most profitable companies in history are finding it hard to cover the bill from income alone.

If they have to borrow, smaller players have far fewer options. That gap is what Nvidia's arrangement is meant to fill.

Turning compute into an asset class

Project finance has long paid for things with big up-front costs and steady returns: power plants, toll roads, airports. The new idea is to treat computing capacity the same way, so that pension funds and insurers are comfortable owning it the way they already own infrastructure.

That means retirement and insurance money would sit on the other side of a bet on AI demand.

Fortune's analysis argues the comparison is imperfect. Energy facilities are built to last decades, usually in response to foreseeable demand, and they generate predictable returns. GPUs, the analysis says, are a different proposition.

The same analysis argues the contest will turn on who can raise capital cheaply and at scale. It says that favors the United States, whose capital markets are deeper than anyone else's, and that China and Europe will have to close a financial gap before they can close the technological one. That is Fortune's argument, not a settled outcome.

What the bank CEOs said this month

The lenders are not shy about the opportunity. They are also not shy about the risk.

During this month's earnings season, Goldman Sachs CEO David Solomon said he was proud of the firm's position in the AI investment cycle. He added a caution on an analyst call: "As the entire AI ecosystem matures, no one knows how demand for capital and compute capacity will evolve." He said that uncertainty is why borrower selection and risk management matter.

Bank of America CEO Brian Moynihan said the massive AI-related capital spending is important for the bank's investment-grade debt and leveraged finance businesses.

Wells Fargo CEO Charlie Scharf said on a post-earnings call that "the AI investment cycle is itself becoming a complex macroeconomic factor." He cited risks including heavily leveraged financing and price pressures in certain industries and supply chains. In his view, economic resilience and the positive effects of the investment currently outweigh the downsides.

Morgan Stanley voiced concern about peers relaxing lending standards, according to reports of the bank's comments. The firm did not, in the accounts of its remarks, name which competitors it meant.

Banks disagree on what AI is doing for them

The banks are also using AI internally, and the results differ. Citigroup, led by CEO Jane Fraser, reported a 40% reduction in development timelines and a 30% cut in external contractor spending. JPMorgan Chase indicated that visible results from its own AI adoption would take time to materialize.

Those are company-reported figures, and neither bank has published the underlying data in the accounts of its remarks.

What the structure leaves open

The Nvidia platforms and the bank lending books rest on the same assumption: that demand for compute will keep pace with the build. Nobody in the accounts of these deals claims to know that it will.

Nvidia's own growth is tied to the answer. Fortune's analysis says the company's expansion depends as much on whether customers can keep affording its products as on engineering.

If demand holds, pension funds and insurers get steady returns and the hyperscalers get their data centers. If it softens, the lenders Morgan Stanley is worried about will be the first to find out. Solomon's question of how demand for capital and compute capacity will evolve has no answer yet.