Updated Microsoft Deal Lets OpenAI Use Google Cloud and Others

OpenAI and Microsoft have updated their partnership with a new amended agreement that changes how OpenAI can deliver its products across cloud platforms. The revised deal preserves a close relationship between the two companies while removing certain exclusivity provisions that previously constrained where OpenAI could host its services.

The collaboration between Microsoft and OpenAI has been a major factor in OpenAI’s rapid growth, helping scale ChatGPT and other offerings quickly. Under the new terms, Microsoft remains an important partner, but OpenAI gains more flexibility to make its products available across multiple cloud providers instead of being limited to Azure for some services.

The key points summarized from the companies’ statements are as follows:

  • Microsoft will remain OpenAI’s primary cloud partner, and OpenAI products will initially be released on Microsoft Azure unless Microsoft cannot or chooses not to support necessary capabilities. OpenAI can now serve all of its products to customers through any cloud provider.
  • Microsoft retains a license to OpenAI intellectual property for models and products through 2032. That license will now be non-exclusive.
  • Microsoft will no longer receive a revenue share from OpenAI.
  • Revenue share payments from OpenAI to Microsoft will continue through 2030, independent of OpenAI’s technological progress, at the same percentage but subject to a total cap.
  • Microsoft remains a major shareholder and will continue to participate directly in OpenAI’s growth.

The most consequential element of the amendment is the elimination of exclusivity for cloud hosting. Previously, certain OpenAI services—including some API access—were available only via Microsoft Azure. With the updated agreement, OpenAI can distribute every product across competing cloud platforms. That change opens the door for customers to access OpenAI services on alternative clouds, which could increase competition among cloud providers and potentially lead to lower costs and better service options for enterprise and developer customers.

Allowing OpenAI to serve all products on any cloud also has strategic implications. Cloud providers that previously had limited or partial roles in hosting OpenAI workloads may now pursue deeper partnerships and infrastructure investments. Greater distribution flexibility helps OpenAI reach customers who prefer particular cloud ecosystems for regulatory, latency, pricing, or architectural reasons. For organizations evaluating where to run AI workloads, the ability to choose a preferred cloud without losing access to OpenAI’s capabilities is a significant shift.

While Microsoft keeps its status as the primary cloud partner and maintains long-term licensing and ownership stakes, the non-exclusive license and the end of a revenue-sharing arrangement mark a notable realignment. The continuation of capped revenue-share payments through 2030 provides some transitional financial structure, but removing exclusivity and revenue sharing changes the commercial dynamics between the two companies and among cloud providers competing for AI customers.

For developers, enterprises, and other customers, the immediate impact may be more choices for deployment and pricing. In practice, this could mean that APIs, hosted models, and managed services from OpenAI become available on a wider range of platforms, allowing customers to integrate OpenAI capabilities into existing cloud-native architectures more seamlessly. Over time, broader cloud availability could encourage more diverse integration patterns, improved redundancy, and potentially lower latency for users by enabling regional or provider-specific hosting.

More on AI:

  • OpenAI continues to iterate on its model lineup, focusing on improvements in context understanding and broader applicability across use cases.
  • Cloud providers and AI companies are increasingly optimizing models and tooling for specific development workflows, such as coding, content generation, and multimodal applications.
  • As AI services become more widely distributed, expect ongoing competition around pricing, performance, and compliance features tailored to enterprise needs.

Follow Ben: Twitter/X, Threads, Bluesky, Instagram