OpenAI locks in $130B nonprofit stake before AGI arrives

OpenAI has finalized a corporate restructuring that puts its nonprofit arm—now called the OpenAI Foundation—in a position of extraordinary financial leverage as the company barrels toward artificial general intelligence. The recapitalization, completed after nearly a year of negotiations with the attorneys general of California and Delaware, gives the Foundation direct equity in the for-profit entity currently pegged at roughly $130 billion. That valuation makes it one of the best-resourced philanthropic organizations in history, and the deal includes provisions for additional ownership as the company hits future milestones.

Sam Altman’s shop isn’t being subtle about the stakes. The whole architecture is designed so that as OpenAI Group PBC—the new public benefit corporation housing the commercial business—grows more valuable, the Foundation’s war chest swells in lockstep. It’s a direct pipeline from commercial success to mission-driven spending, and the Foundation is already earmarking an initial $25 billion across two specific lanes: health research and what it calls “AI resilience.” The health push will fund open-source frontier datasets and scientific grants aimed at diagnostics and cures. The resilience work is framed as a cybersecurity-style layer for AI itself—protecting critical infrastructure from the risks that come with more capable systems.

This isn’t a charity side project bolted onto a tech giant. The Foundation retains control over the for-profit business, a governance arrangement OpenAI claims gives it the strongest mission-focused structure in the industry. The company was founded as a nonprofit in 2015, and despite the 2019 creation of a capped-profit arm that eventually attracted billions from Microsoft and others, the mission—ensuring AGI benefits all of humanity—remains the North Star, at least on paper. The recapitalization formalizes that tension between velocity and safety into a legal framework.

What’s striking is the specificity of the $25 billion commitment at this stage. Most AI safety talk stays in the realm of principles and committees. OpenAI is putting a dollar figure and a defined scope behind the rhetoric, building on the earlier $50 million People-First AI Fund. Whether $25 billion is enough to meaningfully steer a technology this powerful is an open question, but it’s a serious attempt to put real resources behind the governance structure before AGI becomes a reality rather than a research target. The attorneys general review adds a layer of external validation—or at least oversight—to a process that could otherwise look like a company marking its own homework.