Microsoft has revised its deal with OpenAI following the ChatGPT creator’s announcement of a $500 billion joint venture with Oracle, Japan's SoftBank Group, and others. The partnership aims to establish advanced AI data centres across the U.S., marking a significant step in the global race for AI supremacy.
At the White House on Tuesday, former President Donald Trump introduced the project, dubbed "Stargate," emphasizing its role in keeping the U.S. competitive against global rivals like China. Chips from Nvidia are expected to power these ambitious data centres.
Since 2019, Microsoft held exclusive rights to develop OpenAI's computing infrastructure. However, the tech giant recently amended its agreement, allowing OpenAI to collaborate with Oracle on infrastructure projects, mainly for research and model training purposes. This adjustment paves the way for OpenAI to diversify its partnerships while still retaining strong ties with Microsoft.
The Stargate venture is structured as a separate entity, with OpenAI holding equity, governance rights, and operational control. It will feature an independent board, chaired by SoftBank’s CEO, Masayoshi Son, and include additional investors, such as the UAE-based firm MGX. Microsoft, Nvidia, and Arm are listed as technology partners but do not have direct equity stakes in the venture.
While OpenAI’s collaboration with Oracle grows, Microsoft maintains its exclusivity over OpenAI’s API services—key to the company’s revenue stream. Microsoft clarified that Oracle won’t be hosting OpenAI’s primary products. The Redmond-based company also highlighted its ongoing revenue-sharing agreements with OpenAI, ensuring mutual financial benefits until their contract ends in 2030.
Additionally, OpenAI has made a new substantial commitment to Microsoft’s Azure cloud computing platform, which will continue supporting its products and training operations. This reaffirmation underlines Microsoft’s critical role in OpenAI’s future developments despite the latter’s expanded partnerships.