Nvidia Says $100 Billion OpenAI Investment Agreement Is Unfinished

Nvidia’s finance chief clarified on December 2 that an announced letter of intent to provide OpenAI with billions of dollars worth of computing gear has not yet become a binding contract. Speaking at the UBS Global Technology and AI Conference, Chief Financial Officer Colette Kress said the companies "haven't completed a definitive agreement" even as they continue to work together. The original letter of intent envisaged deploying at least 10 gigawatts of Nvidia systems to OpenAI under a multi year arrangement.

Kress reiterated that Nvidia already has roughly $500 billion in booked chip sales through 2026, and she emphasized that any chips sold under a finalized arrangement with OpenAI would be incremental to that backlog. The comment sought to position the prospective deal as additive to an already massive pipeline rather than a reallocation of deliveries that were otherwise committed.

Investors reacted to the clarification with a modest rally in Nvidia stock, which rose about 2.6 percent on the news. Market participants have been parsing the company’s public statements for signs about how large scale AI deployments will be supplied and financed, and whether new commercial structures could create conflicts or concentration risks in the rapidly expanding AI compute ecosystem.

The prospect of a vendor taking an ownership stake or making large scale commitments to a customer has drawn scrutiny. Analysts and investors worry about what have been termed circular deals, where technology suppliers invest in customers that then become major buyers of the suppliers’ gear. Such arrangements raise questions about pricing, preferential access to scarce semiconductors, and the degree to which commercial ties can influence competitive dynamics among cloud providers, AI startups, and chip vendors.

The broader context is a surge in demand for specialized chips required to train and run large AI models. Nvidia has been one of the primary beneficiaries of that trend and has reportedly made other strategic investments in AI companies as it seeks to secure long term demand for its accelerators. Reuters noted those moves when reporting on Kress’s remarks, and market observers say the demand picture remains robust even as buyers negotiate the terms and timing of major deals.

Regulators and corporate governance observers are watching how large scale investments and exclusive supply agreements could shape market structure. For customers, commitments that lock in capacity may ensure access to scarce hardware, but they can also limit flexibility and entrench a small number of suppliers. For Nvidia, clarifying the status of the OpenAI talks helps manage expectations while the company balances immediate backlog obligations against opportunities to capture future demand.

As discussions continue, industry participants and investors will be looking for a definitive agreement or further details that clarify how any deal would be priced, scheduled, and governed, and what it would mean for the allocation of scarce AI compute resources across the market.