Summary

On Feb 27, 2026 Microsoft and OpenAI issued a joint statement reaffirming that their partnership and previously shared terms remain unchanged. Microsoft retains its exclusive license and access to OpenAI models/IP, the commercial revenue‑share arrangement remains intact, and Azure is the exclusive cloud provider for stateless OpenAI APIs. OpenAI may still pursue additional compute elsewhere and collaborate with other cloud providers, but stateless API hosting and first‑party OpenAI products (including Frontier) will continue to run on Azure. The contractual AGI definition and related processes are unchanged.

Key Points

• Partnership terms from prior communications (including Oct 2025) remain in effect.
• Microsoft keeps an exclusive license and access to OpenAI intellectual property.
• Revenue‑share arrangement unchanged; it includes revenue from OpenAI partnerships with other cloud providers.
• Azure is the exclusive cloud provider for stateless OpenAI APIs; stateless API calls from any OpenAI third‑party collaboration will be hosted on Azure.
• OpenAI first‑party products (e.g., Frontier) will continue to be hosted on Azure.
• AGI definition and the process for determining achievement are unchanged.
• OpenAI can commit to additional compute elsewhere (e.g., Stargate) while maintaining the partnership.

Impact for engineers

• API hosting: stateless model inference will be hosted on Azure regardless of third‑party collaborations; expect Azure endpoints and enterprise controls.
• Purchasing: stateless APIs can be bought via Microsoft (Azure) or directly from OpenAI, but traffic/hosting remains on Azure.
• Multi‑cloud integrations should account for Azure‑hosted stateless calls; plan network, security, and compliance accordingly.
• Product teams: OpenAI first‑party services will remain on Azure; no change to IP or revenue expectations that would affect licensing or integration models.