icrosoft and OpenAI have revised their multibillion-dollar partnership, introducing structural changes that redefine cloud exclusivity, intellectual property rights, and revenue-sharing terms through the next decade. The amended agreement keeps Microsoft as OpenAI’s primary cloud provider while loosening exclusivity constraints, allowing OpenAI to distribute its products across multiple cloud platforms.
Under the new terms, OpenAI will continue to launch its services first on Microsoft Azure, unless Microsoft declines to support specific requirements. At the same time, OpenAI gains the ability to deploy its models and services across other cloud environments, signaling a shift toward a more hybrid and multi-cloud commercial strategy. Microsoft retains access to OpenAI’s intellectual property through 2032, but the license now becomes non-exclusive, marking a notable departure from earlier tighter integration.
The financial structure also evolves. Microsoft will no longer pay a revenue share to OpenAI, while OpenAI will continue revenue-sharing payments to Microsoft through 2030 under existing percentage terms, subject to a defined cap. Microsoft remains a major shareholder in OpenAI, maintaining direct exposure to its growth. Both companies emphasized ongoing collaboration across AI infrastructure, including large-scale data center expansion, custom silicon development, and cybersecurity applications.
- OpenAI can now deploy its products across multiple cloud providers, beyond Azure
- Microsoft retains non-exclusive IP licensing rights to OpenAI models through 2032
- Azure remains the primary launch platform for OpenAI services
- Microsoft ends revenue-sharing payments to OpenAI
- OpenAI continues revenue-sharing payments to Microsoft through 2030, with a cap
- Microsoft remains a major equity stakeholder in OpenAI
- Joint efforts continue in AI infrastructure, including data centers and custom silicon
🌐 Analysis: This restructuring reflects a maturing relationship between Microsoft and OpenAI as both scale into trillion-parameter models and gigawatt-class AI infrastructure. The move toward non-exclusive licensing and multi-cloud distribution aligns with broader enterprise demand for flexibility and reduces long-term platform lock-in risk.
