Corporate governance experts who criticized last year’s circular Nvidia-OpenAI deal now have something new to examine: an investment structure that has removed the problematic conflict of interest and replaced it with a straightforward equity arrangement. Nvidia’s $30 billion bet on OpenAI is not just financially significant — it is structurally sound in a way its predecessor was not.
The funding round that this investment is part of will raise approximately $100 billion and value OpenAI at $730 billion. Amazon, SoftBank, and Microsoft are expected alongside Nvidia as investors. The $730 billion valuation is extraordinary — nearly twice Anthropic’s recent figure and just below SpaceX.
The original deal, announced last September, was a governance cautionary tale. Nvidia would invest in OpenAI; OpenAI would use the capital to buy Nvidia chips; the money would cycle back through Nvidia’s order books. Critics immediately identified the circular logic, and when it became clear the deal was never formally binding, it was quietly retired. OpenAI simultaneously confirmed it had been developing chip partnerships with AMD and Broadcom, removing any remaining justification for the old structure.
The new arrangement is unambiguously clean. Nvidia purchases equity in OpenAI. OpenAI receives capital without any conditions on how it is spent. The relationship is that of an investor and a portfolio company, with no supply chain entanglement and no conflict of interest. It is the arrangement that the original deal should have been.
OpenAI’s business challenges do not disappear with the improved investment structure. Market share has fallen. Anthropic is competing effectively for enterprise clients. Cash burn is high. Advertising experiments have attracted backlash. Several investors are hedging publicly. These issues will need to be addressed regardless of how Nvidia structures its investment. But at least the investment itself is now defensible — and that is a meaningful improvement over what came before.