Nvidia is in talks to provide roughly $250 billion in financing guarantees for OpenAI to secure a massive data center campus in southern Ohio. The deal would allow the AI startup to lease a 10-gigawatt facility developed by SoftBank’s energy subsidiary, bypassing OpenAI’s lack of an investment-grade credit rating.
The proposed backstop effectively puts Nvidia’s balance sheet in place of OpenAI’s, allowing lenders to price the debt against the chipmaker’s credit rather than the borrower’s. This maneuver is designed to reassure lenders about the project’s funding for a campus that could cost more than $500 billion when including the necessary silicon.
The Ohio Campus and SoftBank’s Role
The project is centered in Piketon, Ohio, 68 miles (109km) south of Columbus, Ohio. It is being built by SB Energy, a subsidiary of SoftBank, as a public-private partnership. Under this arrangement, the US Department of Energy (DOE) allowed SoftBank to construct what would be the world’s largest AI data centre on leased land. Ohio with 166 centres and another 57 planned has the fourth most operational data centres in the United States after Virginia, Texas and California.
While the $250 billion guarantee covers the lease and construction debt, it does not include the chips themselves. Those components are worth an additional $350 billion, and Nvidia is reportedly in separate talks regarding that financing. The first phase of the facility, around 800 megawatts, is expected to be completed in 2028, providing enough electricity to power 640,000 homes.
The site’s power supply is tied to a separate $33bn US government deal with Japan for a natural gas plant. Because the US government controls the power, Commerce Secretary Howard Lutnick is involved in determining who gains access. While OpenAI has been in advanced talks for several weeks, Anthropic, Microsoft, and Google have also approached Lutnick recently.
The Circular Financing Controversy
The arrangement has drawn sharp criticism from financial analysts and investors who describe it as a “circular” loop: a chip supplier underwriting its customer’s ability to continue buying its own chips. Additionally, OpenAI said it would buy $250 billion of cloud services from Microsoft in 2025.
“Nvidia steps in and provides some funding so they can continue buying Nvidia chips. Where the demand for the whole chain is coming from or where the money is coming from becomes a really big issue.”
— Aleksandar Tomic, associate dean at Boston College
Tomic compares this to the 1999 dot-com bubble, where companies bought products from one another to create an illusion of demand. He argues that OpenAI is essentially using Nvidia’s money to buy Nvidia chips rather than using revenue from customer demand to fund the expansion. Tech commentator Ed Zitron called the move such an insane thing to do on so many levels and about as bearish as it gets, especially considering it is being built by SoftBank.
Investor Michael Burry, known for his 2008 housing market short, echoed these concerns on X, stating, Around and around we go. Nvidia to guarantee $200 billion of ChatGPT’s spending on $NVDA chips.
Burry had added to his Nvidia short position on Friday at $210.28. Following the announcement, Nvidia’s stock fell 4.9 percent in midday trading on Monday.
Strategic Stakes for OpenAI and Nvidia
For OpenAI, the move represents a shift toward owning its compute infrastructure rather than renting it from providers like Microsoft, Amazon, and Oracle. This transition allows the company to dictate capacity rather than negotiate for it. OpenAI’s capital structure is increasingly layered, with eight banks now backing a record $40 billion OpenAI loan from SoftBank.
For Nvidia, the guarantee provides demand certainty. By securing the building’s financing, Nvidia locks in years of chip orders from a customer that would otherwise struggle to find the capital. This scale of guarantee is nearly six times larger than Google’s $44 billion guarantee for other companies’ data center rent. Nvidia has already committed more than $40 billion to AI equity positions, including roughly $30 billion in OpenAI and investments in Ilya Sutskever’s Safe Superintelligence.
The broader industry risk is that AI infrastructure spending—set to exceed $700 billion this year and potentially $3 trillion through 2028—may overshoot actual demand. Research has put Big Tech’s off-balance-sheet AI commitments at around $1.65 trillion, with Meta alone accounting for roughly $420 billion. If returns are competed away faster than the debt is paid off, the high-leverage strategy could falter.
Despite the scale of the talks, no agreement has been signed. Both the site and its financing remain subject to change, leaving the ultimate fate of the Ohio campus uncertain.
Related reading
