OpenAI's Restructuring Shifts How It Balances Profit and Mission
OpenAI's 2025 restructuring changes the company's legal structure in ways that give it greater freedom to pursue commercial contracts, even controversial ones involving defense work. The shift reveals how corporate governance design-not just stated values-shapes whether AI companies prioritize safety or revenue.
Under a deal with California Attorney General Rob Bonta, OpenAI's nonprofit foundation now controls the operating company's board. But the restructuring also fundamentally altered how the nonprofit relates to the for-profit business. Profits can now flow to the foundation for philanthropic use rather than being reinvested in the company itself.
Two Different Governance Models
OpenAI and Anthropic both operate as Delaware public benefit corporations controlled by nonprofits. But their internal structures differ in consequential ways.
Anthropic functions as what legal scholars call a socially oriented for-profit. Its primary purpose is safe AI development, overseen by a Delaware purpose trust that appoints a majority of the board. The trust's role is governance, not profit extraction. Commercial decisions are filtered through the safety mission first.
OpenAI's structure now works differently. It functions as an income-generating for-profit whose surplus flows to a nonprofit parent that funds charitable initiatives. The company can generate and distribute profits while claiming those returns support philanthropic goals like disease research.
How Structure Shapes Decisions
Both companies claim to pursue beneficial AI. But their governance structures create different incentives when facing controversial decisions-like Pentagon contracts.
Anthropic's mission-centered design makes trade-offs sharp. If the company declines a defense contract, that's a clear departure from commercial opportunity. The safety commitment remains the core standard against which decisions are measured.
OpenAI's restructuring created more room to maneuver. Government contracts can be framed as revenue investments that ultimately support charitable purposes through the nonprofit's distributions. The company removed earlier constraints that profits be largely reinvested in the business.
Directors at both companies must balance shareholder interests with public benefit. But courts generally defer to board judgment on how to do that. The question becomes: which governance structure makes it easier for a board to justify commercial expansion?
The Strategic Flexibility Problem
OpenAI's new design aligns investor interests with those of the nonprofit owner. Both benefit when the company produces financial returns. This likely makes the company easier to position for a potential IPO at a reported $1 trillion valuation.
Anthropic faces tighter constraints. Its investors and customers in defense projects must weigh the company's safety commitments against commercial realities. If Anthropic declines certain government work, it loses revenue and strategic position in a market where government demand is rising.
That structural difference matters in an industry defined by enormous capital needs and intense competition. OpenAI embedded commercialization within its governance architecture. Anthropic maintained stronger mission focus but at the cost of reduced strategic flexibility.
What This Means for Legal Professionals
Corporate structure determines how boards interpret mission commitments. "Beneficial AI" is not a precise operational standard. Boards retain considerable discretion in deciding whether a contract advances the mission or simply generates revenue.
The Pentagon dispute shows how governance design influences what options appear available to decision-makers. Legal professionals evaluating AI company governance should examine the underlying organizational structure, not just the stated purpose.
The coming years will test which model proves more sustainable: the socially oriented for-profit or the income-generating for-profit. That choice may shape how the next generation of AI companies balance safety, profit, and national security interests.
For more on how AI governance affects legal work, see AI for Legal.
Your membership also unlocks: