OpenAI Ends Azure Exclusivity, Keeps Microsoft Revenue Share

Context Since Microsoft's $1 billion investment in 2019, OpenAI relied exclusively on Azure for hosting its AI services. The partnership gave Microsoft a license to OpenAI’s intellectual property and models through 2032. Over time, the exclusivity became a focal point for competitors and regulators watching cloud‑AI dynamics. The original 2019 deal included an AGI clause that would have ended exclusivity if OpenAI achieved artificial general intelligence; that provision has been removed, making the revenue share independent of any future AI breakthrough.

Key Facts The amended agreement announced today lets OpenAI deploy its products on any cloud infrastructure. OpenAI will continue to remit 20% of relevant revenue to Microsoft, but that obligation is now subject to an unspecified cap and is only assured until 2030. Microsoft’s license to OpenAI’s IP remains in place through 2032, yet it is now non‑exclusive, allowing other clouds to host the models. The cap on payments is not disclosed, but the guarantee ends in 2030.

What It Means Developers gain flexibility to choose AWS, Google Cloud, or regional providers without losing access to OpenAI’s latest models. Microsoft retains a steady revenue stream and a privileged position as Azure’s primary partner, while OpenAI reduces reliance on a single vendor. The capped revenue share introduces uncertainty for Microsoft’s long‑term earnings from the deal beyond 2030. Cloud‑market competition may intensify as OpenAI’s models become available across multiple platforms.

What to watch next How quickly OpenAI signs multi‑cloud contracts and whether Microsoft adjusts its Azure pricing or incentives to maintain workload share.