Nvidia is scaling back the financial support it had been expected to provide for a massive OpenAI data center project in Ohio, marking a significant change in the financing structure of one of the largest artificial intelligence infrastructure developments announced so far. The chipmaker is now expected to initially guarantee less than $120 billion for the project, a substantial reduction from the $250 billion commitment that had previously been under discussion.
The change reflects growing attention from investors toward the risks involved in financing the enormous infrastructure requirements created by the rapid expansion of artificial intelligence. Nvidia has become one of the biggest beneficiaries of the AI boom through demand for its advanced processors, but its involvement in financing data center projects introduces a different kind of financial exposure. Instead of simply selling hardware to AI companies, Nvidia would effectively become part of the financial structure supporting the infrastructure needed to operate those systems.
The Ohio project is being developed by SB Energy, a subsidiary of SoftBank, and is planned as a 10 gigawatt data center facility for OpenAI. If completed as planned, it would rank among the largest data center projects ever announced. The scale illustrates just how quickly AI infrastructure requirements are growing. Training and operating increasingly powerful AI models require enormous amounts of computing capacity, electricity, cooling systems, networking equipment and physical data center space.
Under the revised arrangement, Nvidia is expected to provide financial backing primarily for the first stage of the Ohio development rather than guaranteeing the financing requirements of the entire project from the outset. The initial guarantee is expected to be below $120 billion, although the precise final structure could still change as Nvidia and OpenAI continue discussions. A deal could be finalized soon, but several elements of the broader project remain under negotiation.
OpenAI is also continuing discussions over a binding lease covering the full 10 gigawatts planned for the Ohio facility. Securing such a lease would be an important step because a project of this scale requires long-term certainty around both computing demand and the financial commitments associated with constructing and operating the site. Large data centers can take years to develop and require significant investment before they begin generating returns.
The decision to reduce Nvidia’s initial guarantee comes amid broader questions about how much financial risk the company should assume as the AI infrastructure market expands. Nvidia has enormous exposure to the growth of artificial intelligence because its processors are central to many of the systems used to train and operate AI models. However, guaranteeing huge infrastructure commitments could expose the company to risks that go beyond its traditional semiconductor business.
For investors, the distinction is important. Nvidia’s core business benefits when companies purchase its chips and related technology. A financing guarantee is different because it can potentially leave Nvidia responsible if another party cannot meet its financial obligations. The larger the commitment, the greater the potential exposure. Reducing the initial guarantee therefore allows Nvidia to remain closely involved in the Ohio project while limiting the amount of risk it assumes during the early phase.
The development also comes shortly after Nvidia announced a partnership with six major financial institutions to create financing platforms aimed at attracting more than $500 billion in third-party capital for AI infrastructure. That initiative highlights a broader shift taking place across the technology industry. The amount of money required to build AI infrastructure is becoming so large that technology companies, banks and other financial institutions are increasingly looking for ways to distribute the cost and risk rather than relying on a single company to finance everything.
This approach could become increasingly important as AI companies compete to build larger computing networks. Demand for advanced processors has surged, but processors alone are not enough to support modern AI systems. Companies also need enormous data centers, power generation capacity, high-speed networking, storage infrastructure and sophisticated cooling systems. The cost of building all of this infrastructure can quickly reach tens or even hundreds of billions of dollars.
For OpenAI, the Ohio development represents another step toward building greater control over the physical infrastructure behind its AI operations. Access to dedicated computing capacity can provide greater certainty as the company develops and deploys increasingly demanding AI models. It could also reduce reliance on infrastructure controlled entirely by outside cloud providers.
At the same time, OpenAI faces questions about the economics of funding such an ambitious expansion. The company has attracted an enormous valuation, reportedly reaching approximately $852 billion, but it remains unprofitable. That creates an unusual situation in which the company is valued at a level comparable to some of the world’s largest technology businesses while simultaneously requiring vast amounts of capital to support its growth.
