Nvidia is reducing the financial support it planned to provide for a major OpenAI data center project in Ohio, but the move may point to something larger than a change in one technology deal. It highlights how financing for artificial intelligence infrastructure is increasingly moving toward private credit and institutional investors.
Nvidia had initially been discussing a guarantee of about $250 billion for the Ohio project, which is being developed by SB Energy, a SoftBank-backed company. The guarantee was intended to help OpenAI secure financing for the data center and related infrastructure. After investors raised concerns about Nvidia taking on too much financial exposure, the proposed support was scaled back to less than $105 billion, according to Reuters.
The project is now expected to have capacity of up to 8 gigawatts, with the first 800 megawatts targeted for 2028. Nvidia will also invest $1.5 billion in SB Energy and remain the site’s exclusive chip provider.
The guarantee does not mean Nvidia is simply paying for the data center. Instead, it is designed to support parts of OpenAI’s lease and power obligations and help protect the value of the site. OpenAI would remain responsible for rent, while the wider financing structure is expected to combine equity with project debt and potentially bonds.
That distinction matters because Nvidia is simultaneously trying to build a different model for funding the AI boom. On August 11, the chipmaker announced partnerships with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to establish independent financing platforms capable of mobilizing more than $500 billion in third-party capital for AI infrastructure.
In simple terms, Nvidia appears to be moving from financing individual AI projects itself toward helping create a market in which banks, private-equity firms and other investors finance data centers and computing capacity.
This is important because AI infrastructure requires enormous amounts of capital. Data centers need not only expensive GPUs, but also land, electricity, cooling systems, networking equipment and long-term operating contracts. Traditional technology financing may not be sufficient for projects that can require billions of dollars and operate for decades.
Nvidia itself describes AI compute as becoming an investable infrastructure asset. Its new financing platforms are intended to give institutional investors a way to participate without Nvidia carrying the entire financial burden.
The approach could eventually help create a distinct market for AI infrastructure debt: loans, bonds and other forms of credit backed by data centers, computing equipment and the future revenue those assets are expected to generate.
That makes Nvidia’s reduced OpenAI guarantee more than a story about the company becoming cautious. It may signal a shift in how the AI expansion is financed. Nvidia is still putting substantial capital behind the infrastructure needed for its chips, but it is increasingly trying to bring Wall Street and private capital into the equation.
The model also carries risks. Investors must determine whether AI data centers will generate enough long-term revenue to justify their enormous construction and financing costs. If demand for computing slows or technology becomes more efficient, some projects could produce lower returns than expected.
