OpenAI completes conversion to Public Benefit Corporation, concentrating roadmap control in the Foundation

The restructuring of OpenAI, finalized in October 2025, turns its former for-profit arm into a Public Benefit Corporation controlled by a foundation that appoints and can replace the entire board. For those who built a business entirely dependent on the API, the question is no longer philosophical: it's about who holds the wheel on the roadmap, pricing, and access to the models.

What Actually Changed in the Structure

OpenAI was founded in 2015 as a nonprofit organization, with the stated mission of ensuring that artificial general intelligence benefits all of humanity. In 2019, in order to scale research and product development, the entity created a for-profit arm, always subordinate to the control of the original nonprofit organization.

On October 28, 2025, according to OpenAI's own official "Our Structure" page, the company completed a restructuring that formalizes this relationship under two new names: the nonprofit became OpenAI Foundation, and the commercial arm became OpenAI Group PBC, a Public Benefit Corporation. Unlike a conventional corporation, a PBC is required by charter to pursue its stated mission and consider the interests of all stakeholders, not just shareholder returns.

In practice, OpenAI is saying it went from "a capped-profit company overseen by a nonprofit" to "a company open to investor capital, but with a legal obligation to balance mission and profit." OpenAI itself describes the goal plainly: to give the Group PBC "the structure to raise capital and attract and retain the talent needed to advance the mission," while keeping what it calls governance more aligned with the sector's mission today.

The Mechanism That Decides Who's in Charge

The point that matters most to those building on top of the API isn't the new name, it's the control mechanism. OpenAI Foundation retains, through special voting and governance rights, the power to appoint all members of the OpenAI Group board and to replace them at any time.

This design concentrates decision-making in a specific way:

• The Safety and Security Committee (SSC), which oversees safety practices across the entire operation, including the commercial Group, remains a Foundation committee, not a Group committee.
• With the exception of Zico Kolter, who chairs the SSC and sits exclusively on the Foundation board (as a full, voting member), all other Foundation directors also sit on the Group board; Kolter is the one who acts as a non-voting observer on the Group board, not on the Foundation board.
• Within one year of the restructuring, a second Foundation director will move to serve exclusively on the Foundation board, also acting as a non-voting observer on the Group board.
• The Foundation's current board includes names such as Bret Taylor (chair), Adam D'Angelo, Paul Christiano, Sue Desmond-Hellmann, General Paul M. Nakasone, Adebayo Ogunlesi, Nicole Seligman, David Vélez, and Robin Vince, along with CEO Sam Altman.

In other words: the entity that decides what is "safe" to release, lock down, or throttle on the API doesn't answer to the market or to the Group's commercial customers; it answers to a board appointed by a foundation whose formal mission can, by design, override commercial logic.

The Number Behind the Move

The restructuring also redefined who gets the biggest slice. According to OpenAI itself, the Foundation holds 26% of OpenAI Group's equity, valued at approximately $130 billion based on the company's current valuation. Doing the simple math from these two numbers, OpenAI Group's implied valuation comes out to around $500 billion, which gives a sense of the scale of the bet.

Beyond its direct stake, the Foundation also holds a warrant: if the Group's per-share value rises more than tenfold within 15 years, it receives a significant additional allotment of shares. This design makes the Foundation, according to the company itself, the largest long-term beneficiary of OpenAI's commercial success, which ties the foundation's cash directly to the growth of the business it also controls.

The process wasn't unilateral: OpenAI says the restructuring was completed after nearly a year of dialogue with the offices of the Attorneys General of California and Delaware, the regulators who oversee this type of corporate conversion involving assets of philanthropic origin.

The Risk for Those Who Bet Everything on the API

This is where it gets relevant to anyone who founded a company on top of OpenAI's API without a plan B. The control architecture that the company itself describes concentrates three decisions in a single board, appointed by a single foundation: what makes it into the model roadmap, how the safety policy handles the release of capabilities (and, by extension, what becomes available via the API), and, indirectly, how the company prices access when it needs to balance mission with cash.

For a founder whose entire product depends on calls to OpenAI's API, this is vendor concentration risk compounded by governance risk: the board that decides whether a feature is released, restricted, or discontinued isn't elected by the market, doesn't answer to corporate customers, and can be replaced at any time by the Foundation. It isn't a board overseeing shareholders; it's a foundation with its own mission overseeing (and able to replace) whoever runs the business that sustains other people's revenue.

The Counterargument That Deserves to Be Taken Seriously

The strongest argument against this reading is that the mission lock-in works in the opposite direction: by requiring the Group to consider stakeholders beyond immediate profit, the PBC structure reduces the chance of erratic commercial decisions, the kind that historically hurt most those who built a business on top of a third-party platform, such as abrupt price changes, model deprecation without notice, or product pivots driven solely by short-term investor pressure.

It's also worth noting that the restructuring itself exists to enable capital raising at greater scale, which, in the optimistic reading, means more cash to keep existing models running and evolving, not less predictability.

This counterargument is real, but it doesn't neutralize the concentration risk: a well-intentioned mission lock is still a mission lock, and it can lock in a direction that the API's customer neither controls nor anticipates, especially on safety decisions that, by design, fall under the Foundation's committee, out of reach of any commercial negotiation with the Group.

What This Changes in Practice

For those deciding how to allocate team and money in Brazil, the takeaway isn't to abandon OpenAI's API, it's to stop treating it as neutral infrastructure. Three practical implications:

1. Map out actual dependency: if a critical product feature depends on a specific model, document the plan B (another provider, an open model, an abstraction layer) before you need it under pressure.
2. Track governance signals, not just the product changelog: board composition, changes to the SSC, and decisions about model access are just as relevant to business continuity as an API pricing update.
3. Don't confuse mission language with a commercial guarantee: a PBC may, by legal obligation, prioritize "stakeholder" interests in a way that doesn't align with the interests of the developer paying per token.

The restructuring solves a real problem for OpenAI, enabling capital at scale to compete in the model race. But it shifts part of the cost of this arrangement onto those outside the board who decided, before October 2025, to bet their entire business model on a single API controlled by a single foundation.

Source: OpenAI, "Our Structure" (official page about the reorganization), https://openai.com/our-structure/