OpenAI’s New Cloud Deal with Microsoft Changes Everything for Amazon, Google, and AI Power

For years, access to OpenAI at enterprise scale came with an unspoken condition that sooner or later, an organization would go through Microsoft.

Azure was not simply a preferred cloud provider; in fact, it was the commercial gateway to OpenAI’s most valuable models, giving Microsoft extraordinary leverage in the race to dominate enterprise AI infrastructure. If businesses wanted serious deployment of GPT models, Microsoft sat firmly in the middle of the relationship.

But right now, a major rewrite of the OpenAI-Microsoft pact has ended what was effectively OpenAI’s cloud exclusivity, allowing the company to distribute its models across multiple cloud platforms (OpenAI, 2026). It sounds like a contractual adjustment, but in practice, it reshapes one of the most important power structures in AI. According to the overhaued pact:

• Microsoft remains OpenAI’s primary cloud partner, and OpenAI products will ship first on Azure, unless Microsoft cannot and chooses not to support the necessary capabilities. OpenAI can now serve all its products to customers across any cloud provider.
• Microsoft will continue to have a license to OpenAI IP for models and products through 2032. Microsoft’s license will now be non-exclusive.
• Microsoft will no longer pay a revenue share to OpenAI.
• Revenue share payments from OpenAI to Microsoft continue through 2030, independent of OpenAI’s technology progress, at the same percentage but subject to a total cap.
• Microsoft continues to participate directly in OpenAI’s growth as a major shareholder.

And Amazon is already first through the door. Amazon CEO Andy Jassy confirmed that OpenAI’s models will soon be available directly through AWS Bedrock, with rollout expected in the coming weeks (Investing.com, 2026). That means companies already built on AWS will no longer need to route critical AI infrastructure through Azure just to access OpenAI’s models.

More importantly, it shows that OpenAI is no longer comfortable being strategically tied to a single platform, even one as powerful as Microsoft.

The original structure made sense when OpenAI needed capital, compute, and stability. Microsoft provided all three, alongside one of the most consequential investments in modern tech. But OpenAI has entered a different phase of its life cycle: scaling.

And scale in enterprise AI is not only about having the best model. It is about distribution.

Most large companies are unwilling to redesign their cloud infrastructure around a single AI provider. They want AI capabilities where they already operate, whether that means AWS, Google Cloud, or another platform. Asking them to migrate core systems to access the model creates friction. Meeting them where they are creates revenue and the revised agreement reflects that reality.

OpenAI can now sell through any cloud provider. Microsoft still retains first access to Azure deployments, preserving a major advantage, but it no longer controls the entire commercial channel. That is a significant strategic shift because infrastructure exclusivity is often more valuable than ownership.

Another major change is the removal of the so-called AGI trigger, one of the strangest clauses in modern corporate dealmaking.

Under the previous agreement, Microsoft’s intellectual property rights were partially tied to OpenAI declaring it had achieved artificial general intelligence. In theory, it sounded visionary. In practice, it created a business dependency around one of the vaguest concepts in technology. Who defines AGI? When does it officially arrive? What happens if both sides disagree?

Basing one of the industry's most important commercial partnerships on a philosophical milestone was always unstable. Both companies eventually recognised that. The revised deal removes the clause entirely and replaces it with something far more practical: Microsoft now holds non-exclusive rights to OpenAI’s product IP, excluding research, through 2032.

That gives Microsoft long-term access without forcing either side into an increasingly awkward debate about whether AGI has been reached.

It also gives OpenAI something equally important: freedom.

Microsoft still walks away with enormous advantages. It no longer has to share Azure revenue generated from serving OpenAI models, strengthening its cloud margins. It continues to receive 20% of OpenAI revenue until 2030, although that figure is now subject to a cap the companies have not publicly disclosed. OpenAI must still ship models to Azure first, preserving Microsoft’s speed advantage, and Microsoft keeps valuable product IP rights well into the next decade.

But what it loses is exclusivity: owning the only bridge to OpenAI created a very different kind of power than being one of several major partners. Microsoft remains powerful, but it is no longer the only road in.

That creates immediate opportunities for Amazon and potentially for Google.

AWS benefits first because it can now offer OpenAI models directly to Bedrock customers without forcing them into a Microsoft ecosystem. That strengthens Amazon’s enterprise AI position at exactly the moment cloud providers are fighting to define the next decade of infrastructure.

Google is watching closely as well. Reports suggest the company is already reviewing the revised terms to explore what kind of partnership may now be possible. If OpenAI models eventually expand across both AWS and Google Cloud, Microsoft’s moat becomes significantly narrower.

At that point, this stops being a model war and becomes something much larger: a distribution war. OpenAI needs enterprise revenue, and it needs it quickly.

As the company nears an expected IPO later this year, investors will care less about cultural dominance and more about recurring revenue, diversified partnerships, and reduced dependence on a single strategic backer. Selling through multiple clouds solves all three problems at once.

It expands market access, improves negotiating leverage, and reassures investors that OpenAI is building a business, not remaining a brilliant lab attached to Microsoft’s balance sheet.