OpenAI and Microsoft Sign New Agreement, Ending Legal Threat From Amazon Deal
Key Points
- Microsoft and OpenAI replace exclusive license with a non‑exclusive deal lasting until 2032.
- Azure stays OpenAI's primary cloud, but the AI maker can now run products on any provider.
- Amazon's up‑to‑$50 billion investment had granted AWS exclusive rights to Frontier and stateful runtime tech.
- The new agreement resolves the legal conflict between Microsoft’s exclusivity clause and Amazon’s deal.
- Microsoft stops paying a revenue share to OpenAI; OpenAI continues a capped share to Microsoft through 2030.
- Microsoft remains a major shareholder in OpenAI, owning about 27 percent of the for‑profit entity.
- Enterprises gain the ability to choose between Azure, AWS, or other clouds for OpenAI services.
Microsoft and OpenAI announced a renegotiated partnership that replaces the exclusive license Microsoft held on OpenAI's models with a non‑exclusive arrangement lasting until 2032. The deal clears a legal snag created by Amazon's up‑to‑$50 billion investment, which had granted AWS exclusive rights to certain OpenAI technologies. Azure remains OpenAI's primary cloud, but the maker can now run its products on any provider. The revised terms also reshape revenue‑share flows and preserve Microsoft’s sizable equity stake, while giving enterprises more cloud‑choice flexibility.
Microsoft and OpenAI unveiled a revised partnership on Monday, swapping the long‑standing exclusive license on OpenAI’s models for a non‑exclusive agreement that runs through 2032. The change removes a legal cloud that had hovered over OpenAI since the AI lab secured a multi‑billion‑dollar investment from Amazon.
Under the new terms, Microsoft retains a non‑exclusive license to OpenAI’s intellectual property for models and products, and continues to be called the company’s “primary cloud partner.” Azure will still host the bulk of OpenAI’s workload for the six‑year span of the contract, even as OpenAI builds its own data‑center capacity with other partners.
Most importantly, the agreement allows OpenAI to offer its products on any cloud provider. Earlier this year, Amazon announced an investment of up to $50 billion—$15 billion upfront and $35 billion contingent on undisclosed conditions. In return, AWS secured exclusive rights to host OpenAI’s new agent‑building tool, Frontier, and to co‑develop a “stateful runtime” technology on AWS Bedrock. Those exclusivity clauses conflicted with Microsoft’s original contract, which gave the software giant exclusive access to OpenAI’s APIs and, potentially, to Frontier itself.
Microsoft had previously warned that it might pursue legal action to enforce its exclusive rights. The revised deal eliminates that conflict, ending the legal peril that could have pitted the two cloud giants against each other. Amazon CEO Andy Jassy celebrated the development on X, noting that OpenAI’s models would soon be available to customers on AWS Bedrock.
Financially, the new arrangement reshapes revenue sharing. Microsoft will no longer pay a revenue share to OpenAI, while OpenAI will continue to remit a capped share of its revenue to Microsoft through 2030. The exact cash flow remains unclear, but Microsoft reported a $7.5 billion quarter‑long boost from its OpenAI investment earlier this year. The tech giant also holds roughly 27 percent of OpenAI’s for‑profit entity, reinforcing its stake in the AI maker’s growth.
For enterprises, the shift delivers greater flexibility. Companies can now select the cloud that best fits their needs while still accessing OpenAI’s models, whether on Azure, AWS, or another provider. The move also underscores a broader industry trend: AI developers are courting multiple cloud partners to avoid lock‑in and to broaden market reach.