Monday, August 3, 2026

Compute & Cloud

Microsoft shifts OpenAI strategy as AI revenue hits $37 billion

Microsoft has revised its OpenAI partnership, ending exclusive access while retaining royalty-free IP rights through 2032 and securing more than $250 billion in future cloud commitments.

Microsoft shifts OpenAI strategy as AI revenue hits $37 billion

On Wednesday, Microsoft reported its earnings for the last full quarter under its previous OpenAI agreement. During the earnings call, a Wall Street analyst asked CEO Satya Nadella how the revised partnership would impact Microsoft’s financials. Nadella asserted that the new agreement is a good deal for everyone, framing it as a win-win construct. Under the updated terms, Microsoft has retained access to OpenAI’s intellectual property—including its models and agent products—but no longer has to pay OpenAI for them. Referring to this royalty-free access through 2032, Nadella stated: “We have a frontier model, with all the IP rights that we will have access to all the way to ’32 and we fully plan to exploit it.”

The changes to the partnership mean Microsoft no longer has exclusive access to OpenAI’s technology. Following the announcement, OpenAI immediately introduced AI products with Amazon, Microsoft’s cloud rival, featuring collaborations between OpenAI executive Sam Altman and AWS CEO Matt Garman.

Despite the end of exclusivity, the financial and operational ties between Microsoft and OpenAI remain deep. The financial relationship is anchored by several key figures:

  • Cloud commitment: OpenAI is committed to buying more than $250 billion worth of Microsoft’s cloud services.
  • Equity stake: Microsoft holds a 27% stake in OpenAI.
  • AI business scale: Microsoft’s AI business has surpassed an annual revenue run rate of $37 billion, representing 123% growth year-over-year.

Nadella noted that OpenAI remains a large customer for Microsoft, purchasing both specialized hardware for AI workloads—which the industry refers to as AI accelerators—and other compute services.

At the same time, Microsoft is positioning itself as a multi-model provider rather than relying solely on OpenAI. Nadella emphasized that Microsoft offers the broadest selection of models of any large-scale cloud service provider (decoding the term hyperscaler), allowing customers to choose the right model for the right workload across OpenAI, Anthropic, and open-source options. According to Nadella, over 10,000 customers have used more than one model. This strategy allows Microsoft to serve diverse enterprise needs while maintaining its equity and cloud revenue streams.

Why it matters

Microsoft is navigating a shift in its OpenAI partnership, moving from exclusive access to a broader model strategy while maintaining significant financial ties and cloud service commitments.