Microsoft and OpenAI End Azure Exclusivity in Partnership Overhaul
Overview
Microsoft and OpenAI announced a major restructuring of their partnership on April 27, 2026, formally ending the Azure exclusivity arrangement that had defined their relationship since 2022. Under the new terms, OpenAI is free to offer its products through any cloud provider, while Microsoft retains first-call status and major shareholder position.
Key Changes
- End of exclusivity: OpenAI can now serve all its products to customers on any cloud platform. Microsoft remains the primary partner; OpenAI products ship first on Azure unless Microsoft cannot support the required capabilities.
- Revenue share: Microsoft will stop paying its revenue share to OpenAI. Payments from OpenAI to Microsoft continue through 2030 at the same percentage, but are now subject to a total cap.
- Licensing: Microsoft's license to OpenAI IP for models and products remains valid through 2032, converting to a non-exclusive arrangement.
- Shareholder status: Microsoft continues to participate directly in OpenAI's growth as a major shareholder.
Background
The restructure was triggered in February 2026, when Amazon committed up to $50 billion to OpenAI with a condition that AWS would have exclusive rights to serve OpenAI's new agent-building tool, Frontier. This directly conflicted with OpenAI's existing Azure exclusivity, forcing a negotiated resolution.
What Comes Next
The day after the announcement, on April 28, OpenAI models became available on Amazon Bedrock, marking the first practical rollout of multi-cloud deployment. OpenAI's platform strategy is now decoupled from any single cloud provider.
Source: Microsoft Blog, OpenAI
Related Articles
Microsoft and OpenAI restructured their partnership on April 27, ending Azure cloud exclusivity and removing the controversial AGI clause. OpenAI can now serve customers on any cloud platform and is free to pursue multi-cloud deals ahead of a potential 2026 IPO.
AWS customers can now access OpenAI's GPT models and Codex coding agent through Amazon Bedrock, marking OpenAI's first major deployment outside Microsoft Azure. General availability is expected within weeks.
This matters because the next bottleneck in agent coding is human attention, not raw model speed. OpenAI says Symphony lifted landed pull requests by 500% on some teams after engineers hit a practical ceiling of roughly three to five concurrent Codex sessions.
Comments (0)
No comments yet. Be the first to comment!