Microsoft and OpenAI End Exclusivity Deal — OpenAI Can Now Work With Any Cloud Provider

By Hector Herrera | April 27, 2026 | Business

Microsoft and OpenAI have amended their partnership agreement, ending Microsoft's exclusive rights to OpenAI's models and freeing the AI company to deploy its products across any cloud provider — including direct rivals AWS, Google Cloud, and Oracle. The deal also removes the contested AGI clauses that had loomed over the partnership for years, resolving two of the biggest structural tensions in the AI industry's most closely watched business relationship.

Background

Microsoft has invested roughly $13 billion in OpenAI since 2019. In exchange, it received a revenue share from OpenAI and exclusive rights to distribute OpenAI's models through Azure. That arrangement gave Microsoft a meaningful competitive advantage in the enterprise cloud market — and placed OpenAI in an awkward position as it grew into a standalone enterprise software company. The original agreement also included AGI provisions: contractual language that could have allowed OpenAI to declare it had achieved artificial general intelligence (AGI — AI that matches or exceeds human reasoning across tasks) and use that declaration to fundamentally alter or end the partnership.

What Changed

According to CNBC, the amended deal makes four significant changes:

• Microsoft's exclusivity ends. OpenAI can now sell and deploy its models on AWS, Google Cloud, Oracle, or any other cloud platform.
• Microsoft stops paying a revenue share to OpenAI. That payment direction is eliminated.
• OpenAI's payments to Microsoft continue through 2030, but under a new cap structure that limits what OpenAI owes as its revenue scales.
• The AGI clause is removed. OpenAI can no longer use an AGI declaration to trigger a partnership restructuring.

Impact

For OpenAI, this is a commercial unlock. The company has been signing enterprise deals with customers who operate primarily on AWS or Google Cloud. The exclusivity clause was a hard ceiling on that growth — customers wanting to use OpenAI's models often had to consider migrating workloads to Azure first. That friction is now gone. OpenAI can pursue cloud-agnostic enterprise contracts without asking customers to change infrastructure.

For Microsoft, the trade is real but manageable. Azure loses its monopoly on OpenAI compute. But Microsoft's investment has already generated returns through GitHub Copilot, Microsoft 365 Copilot, and Azure enterprise deals built on top of OpenAI's models. Microsoft has also diversified its AI bets — through investments in other AI companies and its own Phi model series — reducing its dependence on OpenAI as a single supplier.

For AWS and Google Cloud, this is a direct opportunity. Both platforms have been building OpenAI integrations in anticipation of this kind of opening. Enterprise customers who preferred those platforms but wanted OpenAI's models now have a clear path.

For the broader market, the barrier to enterprise AI adoption drops. Multi-cloud flexibility has been a consistent ask from large organizations. The exclusivity clause was quietly blocking deals.

What to Watch

Whether OpenAI signs formal infrastructure agreements with AWS or Google Cloud — and on what pricing terms — will show how fast it moves to diversify away from Azure. The removal of the AGI clause is also significant: it suggests both parties now view AGI as a research milestone rather than a legally meaningful trigger. Watch for how OpenAI's enterprise revenue mix shifts across cloud providers over the next several quarters.