Nvidia has scaled back its proposed financial guarantee for OpenAI's planned 10-gigawatt data center in Ohio, responding to investor concerns about risk exposure and the unprecedented scale of the project's financing requirements
Nvidia has revised its proposed financial commitment to OpenAI's planned data center campus in Pike County, Ohio, reducing the scale of its guarantee from an earlier figure of approximately $250 billion to under $120 billion. The adjustment follows heightened scrutiny from investors over the financial risks associated with backing one of the largest artificial intelligence infrastructure projects ever proposed in the United States. The guarantee, if finalized, would support financing arranged by external lenders rather than representing a direct investment by Nvidia. This structure is intended to help OpenAI, which does not hold an investment-grade credit rating, secure the substantial debt required for the project's construction and operation.
The Ohio campus is designed to deliver up to 10 gigawatts (GW) of computing power at full build-out, a scale that would place it among the world's largest data center developments. SB Energy, a subsidiary of SoftBank, is leading the site's development. The first phase is expected to provide around 800 megawatts (MW) of capacity by 2028, with subsequent expansion dependent on additional construction and power infrastructure. Achieving the full 10 GW target would require extensive upgrades to electrical systems, advanced cooling solutions, and dense computing hardware. At maximum scale, the facility could house hundreds of thousands of advanced AI processors, supporting the training and deployment of large language models and other generative AI systems.
Nvidia's proposed guarantee is separate from a parallel arrangement under discussion that could see the company supply up to $350 billion worth of AI chips to the Ohio facility. While the guarantee would help OpenAI access debt financing, the chip supply agreement would involve Nvidia providing hardware rather than direct capital. The distinction between these commitments is significant, as each carries different financial and operational implications for Nvidia and its partners. The company's willingness to back such large-scale infrastructure reflects its central role in the current AI hardware ecosystem, but also exposes it to new forms of financial and operational risk.
Investor concerns have focused on the scale of Nvidia's potential exposure, especially as the company's financial relationships with major AI developers deepen. The original $250 billion guarantee proposal raised questions about the prudence of assuming such risk for a project that could ultimately cost more than $500 billion if completed as planned. The revised guarantee is intended to limit Nvidia's liability while still enabling OpenAI to pursue its ambitious infrastructure goals. The terms of both the guarantee and the chip supply arrangement remain under negotiation, and no final agreement has been announced. The Ohio project's progress will depend not only on financing but also on the availability of sufficient electrical power and the ability to deliver the necessary construction and engineering capacity.
The scale and complexity of the Ohio data center highlight the broader challenges facing the AI infrastructure sector. Companies are now planning facilities that require unprecedented capital commitments and long lead times before generating returns. The financing structure-how risk is distributed among technology suppliers, developers, and lenders-has become as critical as the underlying engineering. Nvidia's evolving approach to project guarantees reflects the tension between enabling rapid AI expansion and managing the financial risks of supporting infrastructure at this scale. For context, Nvidia's recent release of Nemotron 3.5 Lightning, an open-weight model designed for high-volume agent workloads, illustrates the company's broader strategy to support both software and hardware advances in AI; more details on that system's deployment can be found in this recent coverage of Nvidia's model releases.
Understanding the distinction between direct investment and financial guarantees is essential in evaluating technology infrastructure projects of this magnitude. A financial guarantee typically means that a company like Nvidia promises to cover certain obligations if the primary borrower-here, OpenAI-cannot meet them. This can make lenders more willing to provide large loans, but it also exposes the guarantor to potential losses if the project underperforms or fails. In the context of AI infrastructure, such guarantees are becoming more common as developers seek to build facilities that far exceed the scale of traditional data centers, raising new questions about risk, accountability, and the long-term sustainability of rapid AI-driven expansion.
