Microsoft was once the exclusive provider of data center infrastructure for OpenAI to train and run its AI models, but that exclusivity has ended. Coinciding with the announcement of Stargate, OpenAI’s massive new AI infrastructure deal with SoftBank, Oracle, and other partners, Microsoft has revealed a new agreement with OpenAI.
This agreement grants Microsoft the “right of first refusal” on new OpenAI cloud computing capacity. Essentially, this means that Microsoft will have the first opportunity to host OpenAI’s AI workloads in the cloud. However, if Microsoft is unable to meet OpenAI’s needs, OpenAI is free to seek services from rival cloud providers.
Microsoft stated in a blog post;
OpenAI recently made a new, large Azure commitment that will continue to support all OpenAI products as well as training. To further support OpenAI, Microsoft has approved OpenAI’s ability to build additional capacity, primarily for research and training of models.
OpenAI has previously cited a lack of available compute resources as a reason for delays in its product development. This shortage of compute capacity has reportedly been a source of tension between OpenAI and Microsoft, which is both a close collaborator and a major investor in OpenAI. In June, under pressure from its shareholders, Microsoft allowed OpenAI to sign a deal with Oracle for additional computing capacity.
In the blog post, Microsoft reiterated that “key elements” of its long-standing partnership with OpenAI will remain in place through 2030. These elements include access to OpenAI’s intellectual property, revenue-sharing arrangements, and exclusivity on OpenAI’s APIs.
However, this agreement assumes that OpenAI does not achieve artificial general intelligence (AGI) under the definition agreed upon by the two companies before 2030. According to a reported agreement, if OpenAI develops AI systems capable of generating at least $100 billion in profits, Microsoft will lose access to OpenAI’s technology. There are indications that OpenAI might be considering nullifying the agreement in an effort to secure more funding from Microsoft.
The blog post reads;
The OpenAI API is exclusive to Azure, runs on Azure, and is also available through the Azure OpenAI Service. This agreement means customers benefit from having access to leading models on Microsoft platforms and directly from OpenAI.