In a sweeping move that could redefine the future of artificial intelligence, OpenAI and Microsoft on October 28, 2025, unveiled a comprehensive restructuring of their partnership, business structure, and intellectual property arrangements. The new agreement, announced jointly and confirmed by statements from both companies as well as regulatory authorities, marks a pivotal moment for the two tech giants and for the broader AI landscape.
OpenAI, famed creator of ChatGPT, has officially converted its business into a public benefit corporation (PBC), following nearly a year of negotiations and regulatory scrutiny. The transition, which also involved a recapitalization, was greenlit by the attorneys general of Delaware and California. In separate statements, Delaware Attorney General Kathy Jennings and California Attorney General Rob Bonta confirmed they would not oppose the restructuring. Bonta emphasized, “We will be keeping a close eye on OpenAI to ensure ongoing adherence to its charitable mission and the protection of the safety of all Californians.” According to AP News, this brings to an end a year-long debate over OpenAI’s governance and the balance of power between its nonprofit board and for-profit investors.
The restructuring paves the way for OpenAI to more effectively commercialize its artificial intelligence technologies while remaining, at least in theory, under the stewardship of a nonprofit board. As Bret Taylor, chair of OpenAI's board, explained in a company blog post, “OpenAI has completed its recapitalization, simplifying its corporate structure. The nonprofit remains in control of the for-profit, and now has a direct path to major resources before AGI arrives.” AGI, or artificial general intelligence, is defined by OpenAI as “highly autonomous systems that outperform humans at most economically valuable work.”
Central to the new arrangement is OpenAI’s deepened partnership with Microsoft. Under the terms of the new definitive agreement, Microsoft now holds a roughly 27% stake in OpenAI Group PBC, valued at approximately $135 billion. This represents a dilution from Microsoft’s previous 32.5% stake, reflecting the impact of recent funding rounds. The agreement preserves key elements from earlier collaborations—OpenAI remains Microsoft’s frontier model partner, and Microsoft retains exclusive rights to OpenAI’s intellectual property and Azure API exclusivity until AGI is achieved. Notably, Microsoft’s IP rights for both models and products are now extended through 2032, even covering models developed after AGI is declared, provided certain safety guardrails are in place.
The technical and commercial modifications are substantial. Microsoft’s rights to confidential research methods—dubbed “research IP”—will remain until either an independent expert panel verifies OpenAI’s AGI declaration or through 2030, whichever comes first. These rights encompass confidential development processes but exclude model architecture, weights, inference code, finetuning code, and data center hardware/software IP. In a departure from previous deals, consumer hardware developed by OpenAI is now explicitly excluded from Microsoft’s IP rights.
OpenAI’s newfound flexibility extends to joint product development with third parties. API products created through such collaborations will remain exclusive to Azure, while non-API products may be served on any cloud provider. This opens the door for OpenAI to diversify its partnerships and infrastructure choices, a significant departure from Microsoft’s prior right of first refusal as OpenAI’s compute provider. In fact, OpenAI has contracted to purchase an additional $250 billion in Azure services, one of the largest cloud contracts in tech history, but is no longer bound to use Microsoft exclusively for computing power.
Another major governance change is the establishment of an independent expert panel to verify any AGI declaration made by OpenAI. Previously, OpenAI’s nonprofit board had unilateral authority to determine when AGI was achieved, a decision that would trigger significant contractual changes. Now, verification by external experts is required, introducing a new layer of oversight and transparency. Microsoft’s rights to OpenAI’s confidential research will remain until either the panel verifies AGI or through 2030. Microsoft will also retain some commercial rights to OpenAI products post-AGI and through 2032.
The nonprofit OpenAI Foundation, which remains in control of the for-profit entity, has committed to granting $25 billion toward health, curing diseases, and protecting against AI cybersecurity risks. However, the timeline for these grants remains unspecified. Importantly, the nonprofit board will soon include at least two members who do not serve on the PBC’s board, addressing concerns about potential conflicts of interest. A Safety and Security Committee within the board retains the power to oversee, review, and even halt new product releases—a safeguard highlighted by the Delaware attorney general’s office.
Microsoft, for its part, can now independently pursue AGI development alone or with third parties, a right that could shift competitive dynamics in the AI sector. If Microsoft uses OpenAI’s IP to develop AGI before a formal declaration, those models will be subject to compute thresholds significantly larger than current leading models. The revenue-sharing agreement between the companies remains in effect until AGI verification, though payments will be spread over a longer period.
OpenAI is also now permitted to provide API access to U.S. government national security customers regardless of the cloud provider, and can release open weight models that meet specific capability criteria. These moves are expected to broaden OpenAI’s reach in both government and commercial sectors.
The origins of this transformation trace back to November 2023, when the nonprofit board’s removal and subsequent reappointment of CEO Sam Altman triggered a reevaluation of OpenAI’s governance structure. The company’s evolution from a nonprofit founded in 2015 to a hybrid model reflects the growing commercial stakes in AI development. Microsoft’s initial $1 billion investment in 2019 laid the groundwork for a partnership that has since grown into one of the most significant alliances in the tech industry.
Despite the regulatory approvals, not all observers are convinced that the nonprofit’s control is more than symbolic. Robert Weissman, co-president of Public Citizen, remarked, “Control is illusory because there is no evidence of the nonprofit ever imposing its values on the for profit.” Meanwhile, OpenAI faces ongoing legal challenges from Elon Musk, a co-founder and early investor, who accuses the company of betraying its original mission. In March 2025, a federal judge denied Musk’s request to block the conversion to a for-profit, but left the door open for an expedited trial on his claims.
Looking ahead, the restructured partnership and business model position OpenAI and Microsoft to continue shaping the future of AI, but not without heightened scrutiny, complex contractual safeguards, and a new era of competition and collaboration. The world will be watching to see how these changes play out as the quest for AGI accelerates.
