Partnerships

Microsoft Loses OpenAI Exclusivity and AGI Clause in Amended Deal

mm

OpenAI and Microsoft amended their long-running partnership on Monday, ending Microsoft’s exclusive license to OpenAI’s models, freeing the AI lab to sell its full product line on any cloud platform, and erasing the controversial “AGI clause” that has shadowed the relationship for years.

The companies announced the changes through coordinated blog posts framed as bringing “long-term clarity” to a deal that has anchored Microsoft’s AI strategy since 2019. Microsoft’s stake — disclosed last October at roughly $135 billion, or about 27 percent of OpenAI on an as-converted diluted basis — remains in place, but nearly every operational lever between the two companies has been retuned.

What Changed in the New Agreement

In Microsoft’s announcement, OpenAI’s products will still ship first on Azure “unless Microsoft cannot and chooses not to support the necessary capabilities,” but OpenAI is now free to “serve all its products to customers across any cloud provider.” Microsoft’s IP license to OpenAI models and products runs through 2032 — but it is now non-exclusive, the most consequential downgrade for Redmond.

The financial structure has been redrawn alongside it. Microsoft will no longer pay revenue share to OpenAI, ending the cut it had handed over for sales of OpenAI models on Azure. OpenAI’s revenue share to Microsoft continues through 2030 “at the same percentage but subject to a total cap,” and the payments are now decoupled from any technology milestone. OpenAI’s matching blog post uses identical language, casting the rewrite as “flexibility, certainty, and a focus on delivering the benefits of AI broadly.”

The most-watched contractual provision is gone. The AGI clause would have triggered changes to Microsoft’s IP rights once OpenAI’s board declared the company had reached artificial general intelligence — an undefined threshold that one party could call unilaterally. The October 2025 “next chapter” amendment tried to defuse it by routing any AGI declaration through an independent expert panel and extending Microsoft’s IP rights to post-AGI models with safety guardrails. This week’s deal removes the trigger entirely; revenue share now stops in 2030 regardless of whether OpenAI declares AGI before then.

Why Both Sides Wanted Out

The amendments codify a multi-cloud reality OpenAI has been building toward for months. The company signed a seven-year, $38 billion AWS deal last November, and Amazon has since committed substantially more in fresh investment. Both moves were hard to reconcile with Microsoft’s prior status as exclusive licensee. OpenAI also continues to push on its own infrastructure even after scaling back the original Stargate expansion in Texas earlier this year, and rival hyperscalers have not slowed down — AWS has been pouring tens of billions into its AI infrastructure and is now positioned to absorb the new OpenAI workloads that Azure exclusivity once blocked.

For Microsoft, the trade-off is losing exclusive access in exchange for stripping out a clause that could have thrown the partnership into chaos at OpenAI’s discretion. Removing the AGI trigger eliminates the single largest source of legal uncertainty in any of the company’s AI contracts, and Redmond keeps the Azure-first deployment posture, the IP license through 2032, and major-shareholder upside as OpenAI continues to grow. Both companies said they will keep collaborating on “scaling gigawatts of new datacenter capacity,” next-generation silicon, and AI-driven cybersecurity — areas where the $250 billion Azure compute commitment OpenAI made in October still binds them tightly.

Markets read the asymmetry. Microsoft shares fell about 2 percent on the announcement while Amazon rose roughly 1 percent, with traders pricing the deal as a clean negative for Redmond’s AI moat and a clean positive for its main cloud rival.

The immediate question is how aggressively OpenAI rolls out its models on AWS and Google Cloud now that the contractual barriers are gone. ChatGPT’s user base and revenue have continued to scale, and a multi-cloud product catalog could land within weeks rather than quarters. For Microsoft, the test is whether losing exclusivity sharpens its push into in-house and partnered models, particularly as Anthropic and other rivals press their own enterprise positions into the same buyers.

The AGI clause’s quiet exit is the deeper signal. For more than half a decade, the most-watched corporate partnership in AI was structured around a vague, undefinable threshold that one side could declare unilaterally. Replacing it with a fixed 2030 sunset on revenue share — and a flat 2032 expiration on Microsoft’s IP license — converts a philosophical bet on the arrival of AGI into a normal commercial contract.

Alex McFarland is an AI journalist and writer exploring the latest developments in artificial intelligence. He has collaborated with numerous AI startups and publications worldwide.