Nvidia is in advanced talks to provide roughly $250 billion in financing guarantees for OpenAI to lease a 10-gigawatt AI data center campus being developed by SoftBank Energy in Piketon, Ohio, according to the Wall Street Journal. The guarantee covers lease and construction financing for the facility but does not include the chips—Nvidia is separately negotiating an additional $350 billion commitment to finance the semiconductor components themselves. The Ohio campus is expected to cost more than $500 billion in total and represent OpenAI's first step toward controlling its own infrastructure instead of renting from Microsoft, Amazon, and Oracle.
The guarantee exists because OpenAI lacks investment-grade credit: the company is projecting losses of approximately $14 billion on roughly $25 billion in revenue in 2026, with internal projections showing cumulative losses around $115 billion through 2029 before reaching cash-flow breakeven. Without Nvidia's balance-sheet backing, no conventional lender would finance the project. The structure converts Nvidia's prior equity investments into a fixed, unhedgeable contingent liability—if OpenAI fails to service the lease debt, Nvidia absorbs the obligation. Commerce Secretary Howard Lutnick controls power allocation for the site; Anthropic, Microsoft, and Google have also approached him about leasing capacity.
For AI infrastructure strategists and financial practitioners, this signals a fundamental shift in AI data center financing architecture: conventional capital markets are deeming standalone AI companies too risky. Nvidia becomes simultaneously chip supplier, financier, and risk absorber—a role that locks in long-term chip orders but carries tail risk. The $250B guarantee is also a public signal that hyperscaler self-sufficiency narratives mask circular financing: Nvidia invests in OpenAI → OpenAI buys Nvidia chips → Nvidia guarantees financing to enable more chip purchases. For investors evaluating AI capex returns, this underscores the concentration of systemic risk in a narrow set of suppliers.