On Tuesday, Microsoft announced adjustments to its agreement with OpenAI after the ChatGPT creator revealed a collaborative effort with Oracle and Japan's SoftBank Group to establish new AI data centers worth up to $500 billion in the United States. This partnership was presented by President Donald Trump at the White House to leverage Nvidia chips in the global AI competition against China and other contenders.
Microsoft had an exclusive deal with OpenAI since 2019 for developing new computing infrastructure. However, they have now authorized OpenAI to expand capacity specifically for research and model training. This modification allowed OpenAI to partner with Oracle in a joint venture known as Stargate, structured as a new entity with its own governance and key investors, including MGX.
The Stargate venture involves Microsoft, Nvidia, and Arm as technology partners, with SoftBank CEO Masayoshi Son appointed as the board chairman. Despite this collaboration, Microsoft maintains its exclusive right to offer OpenAI's API, ensuring that Oracle cannot host OpenAI's primary revenue source.
While Oracle did not immediately comment on Microsoft's statements, Microsoft highlighted revenue-sharing agreements with OpenAI, emphasizing the continuity of essential elements in their partnership until 2030. Additionally, Microsoft mentioned OpenAI's new substantial commitment to Azure, Microsoft's cloud computing service, in supporting all OpenAI products and training.