Back
Bloomberg / Microsoft BlogBusinessBloomberg / Microsoft Blog2026-04-27

Microsoft and OpenAI End Their Exclusive Deal — A $13B Partnership Enters Its Next Phase

Microsoft and OpenAI have restructured their landmark partnership: Microsoft's exclusive license to OpenAI's IP ends, OpenAI can now serve any cloud provider, and Microsoft stops paying its revenue share — while OpenAI continues paying Microsoft through 2030.

Original source

The AI industry's most consequential business relationship just changed shape. Microsoft and OpenAI announced on April 27, 2026 that they are reworking the terms of their original $13 billion partnership in ways that give both companies significantly more independence.

The headline change: Microsoft's license to OpenAI's intellectual property is now **non-exclusive**. OpenAI can sell its models — including GPT-5.5 and future releases — to any cloud provider, including direct rivals Amazon Web Services and Google Cloud. Previously, Azure had first and exclusive access to OpenAI's most capable models.

Financially, the deal cuts both ways. Microsoft will stop paying revenue share to OpenAI, reducing its ongoing obligations under the original arrangement. OpenAI, however, will continue paying a revenue share to Microsoft through 2030, now subject to a total cap — a nod to OpenAI's growing commercial scale and ability to operate without Microsoft's full financial support.

Microsoft remains OpenAI's primary cloud provider, retaining first-access rights on new model releases unless it opts otherwise. The IP license itself continues through 2032. Both companies emphasized that joint infrastructure projects — including AI chip development and cybersecurity collaboration — remain intact.

The market reacted with a 3% drop in Microsoft's share price as investors processed the loss of exclusivity. Analysts were divided: some saw it as Microsoft conceding ground as OpenAI outgrows its original patron; others viewed it as a pragmatic de-risking by Microsoft, which has been expanding its own in-house AI model capabilities through the MAI program launched earlier this month.

For the broader AI ecosystem, the implications are significant. OpenAI models on AWS and Google Cloud creates a new competitive dynamic — Azure is no longer the only enterprise path to GPT. And with Microsoft's own MAI models maturing, the former partners may find themselves competing on capabilities as much as collaborating.

Panel Takes

The Builder

The Builder

Developer Perspective

OpenAI models on AWS and GCP is genuinely good news for developers — we're no longer forced into Azure to access the best models. The competition for inference pricing should push costs down across the board.

The Skeptic

The Skeptic

Reality Check

Microsoft paid $13B and built Copilot on OpenAI's foundation — and now loses exclusivity just as competitors catch up. The 3% stock drop is rational. This is what happens when you don't own the models that power your flagship product.

The Futurist

The Futurist

Big Picture

The decoupling of Microsoft and OpenAI is a structural inflection point. We're entering the phase where frontier AI models are commodities any cloud can resell — and the moat shifts entirely to proprietary data, workflows, and applications built on top.

Bookmarks

Loading bookmarks...

No bookmarks yet

Bookmark tools to save them for later