OpenAI and Microsoft Rewrite Their Alliance Around AGI

Two major technology organisations redefining a shared path while an independent panel evaluates a distant AI milestone

OpenAI and Microsoft have renewed one of the technology industry’s most consequential partnerships while reducing the dependence that originally defined it. The new agreement preserves Microsoft’s privileged role around frontier models, yet gives both companies more freedom to pursue infrastructure, products, and even artificial general intelligence on separate paths.

The deal arrives alongside OpenAI’s corporate reorganisation. Its commercial operation is now OpenAI Group PBC, a public benefit corporation legally tied to the mission. The renamed OpenAI Foundation retains control and holds an equity stake valued at approximately $130 billion.

The structure is intended to turn commercial growth into direct funding for public-interest work, beginning with a $25 billion commitment to global health and AI resilience. OpenAI describes the result as mission-focused governance capable of pushing the technical frontier while directing more of the value toward public benefit.

The Foundation Keeps Control of a Larger Commercial Engine

OpenAI finalised the arrangement after almost a year of discussions with the Attorneys General of California and Delaware and says it changed the structure in response. The Foundation’s ownership means its resources should grow if the PBC becomes more valuable.

That creates an unusual feedback loop: a mission-led nonprofit controls a rapidly expanding commercial company and uses the resulting wealth to fund its goals. The public-benefit status places the mission in the legal design of the business rather than leaving it solely as a promise from executives.

The critical test will be how control works when commercial opportunity and the stated mission diverge. A large equity position provides funding, but it also ties the Foundation’s financial capacity to the success of the company it oversees. Formal control matters most when the profitable choice is not obviously the mission-aligned one.

AGI Now Requires Independent Verification

Microsoft’s investment is valued at $135 billion, representing a 27 percent stake in OpenAI Group PBC, down from 32.5 percent after dilution from later funding. Microsoft remains the exclusive Azure API provider for OpenAI frontier models until AGI is achieved.

OpenAI can no longer end that phase merely by declaring the milestone itself. An independent expert panel must verify any AGI claim. The provision matters because the declaration affects revenue sharing, intellectual property, cloud exclusivity, and the future freedom of both parties.

Microsoft’s IP rights now extend through 2032 and include models developed after verified AGI, subject to safety safeguards. It may also pursue AGI independently or with other partners. If it uses OpenAI IP to reach AGI before an official declaration, substantially higher compute thresholds apply.

Both Companies Gain Room to Manoeuvre

OpenAI has committed to buying an additional $250 billion of Azure services, but Microsoft no longer holds first refusal as its compute provider. OpenAI can therefore negotiate with other infrastructure companies as its demand grows.

OpenAI may release qualifying open-weight models and serve US national-security customers on any cloud. It can jointly develop certain non-API products with third parties, while external API products must remain on Azure. Microsoft’s rights explicitly exclude future OpenAI consumer hardware.

The existing revenue-sharing agreement continues until the expert panel verifies AGI, with payments spread across a longer period. Those details turn a philosophical question—when has AGI arrived?—into a contractual event with enormous economic consequences.

The revised alliance neither dissolves the partnership nor restores its earlier simplicity. Microsoft retains equity, IP, API exclusivity, and a vast Azure commitment. OpenAI gains infrastructure leverage and product flexibility, while Microsoft gains an independent route forward. Their next chapter is built around managed interdependence: close enough to benefit from each other, but no longer willing to make the future of either company depend entirely on the other. Independent AGI verification becomes the hinge holding that balance together.

Exit mobile version