OpenAI Confirms Transition to Public Benefit Corporation, Microsoft Takes 27% Stake as “AI Power Map” Is Redrawn

A Nonprofit Will Directly Oversee For-Profit Operations
Structure Mirrors Earlier Proposals in Broad Terms
Potential Clash Between Public Mission and Commercial Reality

OpenAI has transitioned into a public benefit corporation, declaring a new start under the control of a nonprofit board. With the restructuring, Microsoft (MS) has secured about 27% of OpenAI’s shares, though its control over hardware and cloud infrastructure will be limited. The industry views this reorganization as an expected development but warns that OpenAI’s “public-interest-first” principle could restrict the pace of technological progress and capital inflows. Analysts also note that it could reignite criticism from Elon Musk and other detractors, while tensions with Microsoft may resurface as a new variable.

A Drive for “Technical Independence” Through External Interest Adjustment

On the 28th, OpenAI announced it had signed a new agreement with Microsoft to rebalance capital and proceed with its conversion into a public benefit corporation. Under the new structure, Microsoft will hold about 27% of shares in the newly established “OpenAI Group PBC (Public Benefit Corporation).” Based on valuation, this stake amounts to roughly $135 billion, allowing Microsoft to maintain its position as OpenAI’s largest investor. Meanwhile, OpenAI’s earlier plan to become a fully for-profit corporation has effectively been withdrawn, returning to a structure in which a nonprofit foundation holds control and supervises operations from the top.

The key feature of this governance reform is the “two-tier structure.” The upper entity, the OpenAI Foundation, has been reorganized to hold about $130 billion in voting shares of the PBC. The foundation is also designed to acquire additional shares whenever the company’s valuation surpasses a certain threshold, ensuring that the nonprofit will continue to exert direct control over the for-profit arm. OpenAI Chair Bret Taylor explained that “the goal is to secure control and access to key resources before the arrival of artificial general intelligence (AGI).”

In this process, Microsoft’s 27% stake carries significance beyond a simple financial investment. Under the new agreement, OpenAI granted Microsoft intellectual property (IP) rights to use its models and products until 2032, even in the event its AI models reach AGI capability. However, this right does not apply to consumer hardware developed directly by OpenAI. In other words, Microsoft remains a partner confined to infrastructure—software, algorithms, and model training—without influence over physical products. This reflects OpenAI’s intent to isolate its hardware business, a move widely interpreted as a preemptive measure in preparation for post-AGI market competition.

At the same time, Microsoft lost its priority negotiation rights as a cloud service provider. While OpenAI will continue to use Microsoft’s Azure cloud as its main infrastructure, additional purchase volume will be capped at about $250 billion, and future contracts will be negotiated individually. Their profit-sharing agreement will remain in effect until AGI verification, and Microsoft is estimated to receive around 20% of OpenAI’s profits. The segmentation of IP, equity, and revenue under differing terms and durations reflects OpenAI’s attempt to manage both technological independence and external capital inflows.

OpenAI justified the restructuring as a balance between “public value and the sustainability of the AI industry,” adding that it plans to reinvest a portion of profits into public-interest initiatives while upholding its founding mission of ensuring AGI benefits humanity. The PBC model, which allows external investment while maintaining nonprofit control, is seen as a possible new standard for the AI industry. However, whether this will be perceived as a “declaration of technological independence” or merely as a “formal adjustment to reduce reliance on a specific investor” remains to be seen.

A Compromise Model to Maintain Capital Inflows

Within the industry, many view OpenAI’s restructuring as a predictable outcome. Founded as a nonprofit in 2015, the company added a “capped-profit” for-profit structure in 2019 under nonprofit board control, and by late last year, had begun to move toward a fully for-profit model that could support an independent listing. During that process, OpenAI faced criticism that it prioritized commercialization over its nonprofit mission, along with oversight demands from regulators in California and Delaware, ultimately opting for the middle ground of a public benefit corporation.

In explaining the change, OpenAI stated, “Building AGI requires far more capital than we originally imagined.” The company has already raised several rounds of massive external funding and conducted record-scale private fundraising, bringing in more than $60 billion at a valuation of roughly $157 billion—an unprecedented feat for a private firm. The newly formed OpenAI Group PBC was structured precisely for this purpose. By enabling traditional share issuance, additional rounds, and even a potential IPO, OpenAI gained access to capital avenues that were impossible under its prior capped-profit subsidiary model.

This structure effectively institutionalizes the principle of “control remains with the foundation, funding comes from the market.” Outwardly, it promotes a virtuous cycle of reinvesting AI profits into social impact, yet the for-profit PBC retains the flexibility to grow through additional funding rounds or a public offering. Still, uncertainties persist. If OpenAI fails to reach its self-defined AGI benchmark, or if competitors such as Google, Meta, or xAI narrow the technological gap, its current valuation and funding rationale could face pressure. The company must also sustain regulatory and market confidence to preserve long-term credibility.

Increased Pressure for Transparency and Technology Disclosure

The public-interest mandate may also constrain OpenAI’s operational scope as it expands its technology and business. Because its board is controlled by the nonprofit foundation, OpenAI’s executives are required to prioritize ethical values and social safety over profit. The board also holds unilateral authority to dismiss management if it determines that the company has deviated from its founding principles. OpenAI’s board already exercised this power in November 2023, when it removed CEO Sam Altman, a move that sparked widespread controversy. Maintaining this clause is seen as an effort to strengthen internal checks on the company’s commercialization pace and strategic direction.

For this reason, industry observers view the latest decision as a “halfway outcome” for OpenAI. Axios reported that “CEO Sam Altman sought to free himself from board oversight but has now returned under the supervision of a nonprofit board,” adding, “OpenAI has gained the freedom to raise capital but has also inherited the legal and ethical restraints that come with its public-benefit status.”

These constraints are also expected to embolden OpenAI’s critics. Elon Musk, CEO of xAI, continues to pursue legal action against OpenAI for what he claims is the abandonment of its nonprofit mission. Musk’s legal team argues that “the public-benefit conversion changes form but not substance,” claiming that it represents the continuation of a profit-driven structure under a nonprofit label. The advocacy group Public Citizen similarly criticized the move, stating, “OpenAI’s announcement is business as usual, a way for its board to retain tax advantages under the guise of nonprofit control.” With criticism reigniting, OpenAI now faces heightened scrutiny over transparency, particularly regarding technology disclosure and data access.

Meanwhile, tensions are also rising within its partnership with Microsoft. The Wall Street Journal reported that OpenAI considered filing a complaint with the Federal Trade Commission (FTC) against Microsoft’s excessive control during the restructuring process. Disagreements reportedly centered on the acquisition rights to coding startup Windsurf and the allocation of intellectual property after AGI development. Separately, the FTC is already investigating potential anticompetitive elements in the OpenAI–Microsoft partnership. The Journal noted, “The two companies are now direct competitors in key AI markets such as chatbots and enterprise applications, and the latest restructuring leaves them walking a fine line between collaboration and rivalry.”