OpenAI’s Shift from Nonprofit to Public Benefit Corporation

OpenAI has become a central case study in how society tries to govern powerful technologies while navigating intense commercial pressures. The organisation began as a nonprofit with a mission to ensure artificial intelligence benefits all of humanity, then built a complex “capped-profit” structure, and is now seeking a more conventional for-profit form. This journey raises pressing legal, ethical, and strategic questions about how to align advanced AI with the public interest.

From Idealistic Nonprofit to AI Powerhouse

When OpenAI launched in 2015, it presented itself as a research-oriented nonprofit focused on ensuring that artificial intelligence was developed safely and broadly shared. According to legal analyses of the organisation, its original charter centred on funding and conducting AI research and distributing resulting technologies in ways that would benefit the public rather than maximise private profit.

Several factors made the nonprofit model appealing in the early years:

• Talent recruitment based on values: Being a nonprofit signalled that the organisation prioritised safety and social impact over short-term profit, which reportedly attracted researchers who were uneasy about working for large tech firms whose incentives might be more commercial.
• Philanthropic funding: Initial funding came largely from major donors who were willing to support ambitious AI research as a public good rather than demanding immediate financial returns.
• Public trust: A nonprofit status helped convey a commitment to serving humanity, underscoring the organisation’s stated goal of ensuring that advanced AI would be shared for the benefit of all.

However, AI research at the frontier is extremely capital intensive. Training state-of-the-art models requires large-scale compute, specialised hardware, and highly paid technical talent; estimates from major labs suggest that training costs for cutting-edge systems can reach tens or even hundreds of millions of dollars per model. As the ambition and scale of OpenAI’s projects grew, philanthropic donations alone became insufficient.

Why a Pure Nonprofit Model Struggled

By around 2019, OpenAI’s leaders concluded they could not raise enough philanthropic capital to remain competitive in advanced AI research. This is consistent with broader trends: governments and large technology companies have been the main funders of frontier AI, while philanthropic funding for fundamental AI safety and governance has remained comparatively small and fragmented.

Key pressures that strained the pure nonprofit model included:

• Escalating compute costs: Training larger models requires exponentially more computation and specialised infrastructure, often sourced from major cloud providers.
• Rising salaries for AI experts: Highly skilled machine learning researchers can command compensation packages that rival those at major technology firms, making nonprofit pay scales difficult to sustain.
• R&D race dynamics: As other firms accelerated AI development, staying at the frontier increasingly meant competing in a quasi-arms race for talent, data, and compute.

Faced with these realities, OpenAI sought a structure that would preserve a public-interest mission while permitting large-scale private investment.

The Hybrid Structure: Capped Profit Meets Charitable Control

In response, OpenAI developed a complicated hybrid structure often described as a “capped-profit” model. At a high level, the idea was to house commercial activities in a for-profit entity whose upside for investors would be limited, while maintaining ultimate control in a nonprofit board dedicated to the public mission.

According to nonprofit governance experts, the structure involved several layers:

• A nonprofit parent with a charitable purpose and a board supposedly empowered to ensure alignment with the mission.
• A limited partnership and associated for-profit entities that could raise external capital.
• Contractual and governance mechanisms intended to bind for-profit activities to the nonprofit’s mission, including caps on investor returns and mission-focused charter language.

On paper, this arrangement attempted to reconcile two competing realities:

• The need for large-scale, flexible funding typical of high-growth technology companies.
• The desire to maintain mission-first governance through nonprofit control and fiduciary duties focused on public benefit.

Hybrid structures of this kind are not unprecedented. Legal scholars note that many charities have for-profit subsidiaries, particularly when they engage in substantial commercial activities that might not fit cleanly within standard nonprofit constraints. However, in most cases the nonprofit’s dominance is clear: it owns the subsidiary, appoints its board, and controls its strategy. OpenAI’s situation is more contested because of the immense commercial value created and the evolving balance of power between mission and money.

Board Duties and the Question of Control

The OpenAI case spotlights a core issue in nonprofit law: when a charity controls a subsidiary, its board is expected to ensure that the subsidiary’s activities remain consistent with the charity’s mission and legal obligations. Under U.S. state law, charity directors generally owe duties of care and loyalty to the organisation’s charitable purpose, not to investors.

In practice, however, enforcing this principle in a high-stakes, multi-entity structure can be difficult. Several tensions arise:

• Information asymmetry: Management teams inside the commercial arm may have more detailed knowledge about operations and risks than the nonprofit board, making oversight harder.
• Cultural and incentive drift: As more revenue and valuation accumulate in the for-profit arm, internal decision-making may start to resemble that of a typical tech company, even if formal documents still emphasise public benefit.
• Investor expectations: Sophisticated investors are accustomed to governance structures that prioritise shareholder value. Even when accepting a return cap, they may exert pressure for more conventional corporate arrangements over time.

Episodes such as leadership changes and governance disputes have raised questions about whether the nonprofit board can realistically exercise strong, independent oversight in the face of these pressures. Some analysts argue that the original design underestimated how hard it would be for a charitable board to maintain control over a rapidly growing, commercially valuable AI company.

Proposed Shift to a For-Profit Public Benefit Corporation

More recently, OpenAI has pursued a restructuring that would move away from nonprofit control and toward a more standard corporate architecture, such as a public benefit corporation (PBC). A PBC is a for-profit corporation that includes a specific public benefit purpose in its charter and requires directors to balance shareholder interests with that broader purpose.

Several U.S. states have statutes enabling benefit corporations or similar entities, allowing companies to embed social or environmental missions into their governance documents alongside profit motives. These forms have been used by firms in sectors such as sustainability and impact investing, and some large technology companies have explored mission-driven legal structures for particular initiatives.

Proponents of a transition to a PBC or similar structure argue that:

• Governance becomes clearer: Instead of a convoluted nonprofit–for-profit hierarchy, there is a single corporate form with defined duties.
• Capital raising is easier: Investors are more comfortable with familiar corporate frameworks, especially if the company may eventually go public.
• Mission language is preserved: A public benefit purpose can still be written into the charter, at least symbolically constraining purely profit-maximising behaviour.

Critics worry, however, that this shift could weaken the original commitment to prioritising global benefit over commercial success, especially in a field where the externalities of failure—such as misuse, disinformation, or systemic risks—could be extremely severe.

Mission vs. Market: Deeper Ethical Questions

OpenAI’s changing structure highlights a broader dilemma: how can societies govern technologies with potentially transformative impacts when they are primarily developed by private firms? Concerns raised by researchers, policymakers, and ethicists include:

Governments have begun to respond. For example, the European Union has adopted a comprehensive AI Act that introduces risk-based regulation for AI systems, including transparency, safety, and oversight requirements. In the United States and other jurisdictions, policymakers are exploring voluntary codes, sectoral rules, and funding for public-interest AI research. Nonetheless, much of the day-to-day governance of frontier AI still rests with corporate boards and executives.

Lessons for Nonprofits Working With High-Risk Technologies

OpenAI’s experience offers several lessons for charities and philanthropies that seek to advance or govern powerful technologies:

• Be realistic about capital needs: If delivering on a mission requires large-scale, sustained investment, leaders must decide early whether a pure nonprofit structure is viable or whether hybrid or for-profit forms will ultimately be necessary.
• Design governance for stress conditions: Structures that work in calm periods may fail when valuations soar or crises erupt. Boards should anticipate conflicts between mission and money and embed clear decision rights, escalation paths, and accountability mechanisms.
• Maintain clarity of fiduciary duties: Directors of charities must understand that their primary legal obligation is to the mission and the public interest, even when working with or overseeing commercial subsidiaries.
• Use transparency to build trust: Complex arrangements are especially vulnerable to suspicion. Public disclosure about governance, risk management, and philanthropic commitments can help maintain legitimacy.

At the same time, philanthropic actors and nonprofits may choose alternative strategies, such as funding open-source research, supporting public-sector AI labs, or investing in regulatory capacity, rather than directly building commercial-scale AI systems themselves.

What This Means for the Future of AI Governance

The debate over OpenAI’s structure is not only about one company. It is an early test of how the law, philanthropy, and markets will interact as AI systems become more capable and pervasive. Several open questions remain:

• Can hybrid or public benefit structures meaningfully constrain profit-driven incentives, or do they mostly serve as reputational tools?
• What role should state attorneys general, charity regulators, and competition authorities play in overseeing entities that combine nonprofit missions with massive commercial value?
• How can global governance arrangements ensure that AI development benefits people across countries and generations, not just a small number of firms and investors?

Legal scholars like Rose Chan Loui and others argue that, at least in principle, charity law provides mechanisms for enforcing mission commitments when a nonprofit controls subsidiaries, including regulatory interventions and litigation by state officials. Whether these tools will be deployed in AI-related cases—and whether they will be sufficient in the face of geopolitical and commercial pressures—remains to be seen.

Frequently Asked Questions (FAQs)

Q: Why did OpenAI start as a nonprofit instead of a normal tech company?

A: According to legal and governance analyses, nonprofit status helped OpenAI signal a commitment to public benefit, attract researchers motivated by safety and ethics, and seek philanthropic funding for high-risk research, rather than being driven solely by investor returns.

Q: What is a “capped-profit” model in this context?

A: In OpenAI’s case, a capped-profit model involves allowing outside investors and employees to earn returns from a for-profit entity, but placing an upper limit on those returns and formally subordinating commercial goals to a charitable mission controlled by a nonprofit board.

Q: How is a public benefit corporation different from a regular corporation?

A: A public benefit corporation is a for-profit entity that incorporates a specific public benefit purpose into its charter and requires directors to consider that purpose alongside shareholder value. Traditional corporations generally prioritize shareholder interests by default under corporate law, though the exact interpretation varies by jurisdiction.

Q: Do nonprofit boards have legal power over their for-profit subsidiaries?

A: When a charity controls a subsidiary, its board is expected under state charity law to ensure that the subsidiary’s activities serve the charitable mission and do not improperly benefit private parties. Regulators such as state attorneys general can, in principle, intervene if a charity’s control is undermined or misused.

Q: What role do governments play in governing frontier AI beyond company structures?

A: Governments can create binding rules for AI safety, transparency, and accountability—such as the EU AI Act’s risk-based regulation—and fund public-interest research, standard-setting, and oversight institutions. However, much practical control still lies with private labs and their boards.

Written by

Sneha Tete

• Integrated M.A. in Applied Linguistics, University of Hyderabad
• 5+ years in editorial, instructional, SEO, and research-based writing

Sneha is a relationships and lifestyle writer with a strong foundation in applied linguistics and certified training in relationship coaching. She brings over five years of writing experience to theskillvault,  crafting thoughtful, research-driven content that empowers readers to build healthier relationships, boost emotional well-being, and embrace holistic living. Her writing and insights reflect the belief that beauty, happiness, and wellness…