OpenAI Foundation, the nonprofit that controls the artificial intelligence company behind ChatGPT, pledged Tuesday to grant $1 billion over the next year toward life science research, health initiatives, and programs addressing AI's effects on jobs and children's mental health. The number is real. The context is what makes it worth reading.
The OpenAI Foundation controls assets worth roughly $130 billion, making it the largest nonprofit organization in the United States by valuation. Yet in 2024, the most recent year it filed with the Internal Revenue Service, the foundation received $4,433 in contributions and granted $7.6 million to outside organizations, according to AP News. Its expenses in 2018, before the for-profit subsidiary was incorporated, were $51 million. By 2019 they had fallen to $3.3 million.
The new $1 billion pledge arrives as OpenAI has moved aggressively to formalize its unusual governance structure. In October 2025, OpenAI finalized a restructuring agreement with regulators in California and Delaware that left the nonprofit board in formal control of the for-profit business while allowing investors to profit from its technologies. The $130 billion valuation stems from the nonprofit's 26 percent equity stake in the commercial entity. OpenAI is now valued at $730 billion following a $110 billion funding round in February 2026, per reporting by CNBC.
One detail that nonprofit watchers have flagged as critical: the OpenAI Foundation is classified as a public charity under federal tax law, not a private foundation. That distinction matters. Private foundations are generally required to distribute at least 5 percent of their assets annually. Public charities face no such mandate, according to The Chronicle of Philanthropy. OpenAI is not legally required to spend down its assets on a predictable schedule.
The governance structure compounds the scrutiny. The foundation shares all but one board member with the for-profit company, The Chronicle of Philanthropy reported. It serves as the nonprofit parent, retains formal governance authority, and its fortunes rise and fall with a single company. That has nonprofit leaders and philanthropy experts asking hard questions about independence.
"Is this a true foundation or is it just a drawer on Sam Altman's desk?" said Orson Aguilar, CEO of LatinoProsperity and a member of the Eyes on OpenAI coalition, a group of more than 60 nonprofits that formed in 2025 when OpenAI was actively exploring ways to reduce nonprofit control. The question is not rhetorical. Much of the foundation's $130 billion valuation exists on paper from its equity stake, not as cash available for grantmaking.
Don Howard, president of the James Irvine Foundation, a California-based grant maker focused on economic opportunity for low-income workers, was more blunt. He called OpenAI's initial grants "an embarrassingly small amount of money to commit to nonprofits" given its resources, per The Chronicle of Philanthropy.
Catherine Bracy, who has covered the OpenAI nonprofit debate extensively, put the structural problem plainly: the foundation is never going to make a decision that harms the company. These two entities cannot live under the same roof where the mission is in control, she told Vox. That is the governance question that the $1 billion pledge lands inside, not outside.
The foundation has begun demonstrating activity. In December 2025 it announced $40.5 million in grants through its People-First AI Fund, supporting AI literacy, civic engagement, and workforce development programs. Several of those grantees were members of Eyes on OpenAI, per Vox. The new pledge adds life science and health research as priority areas and includes the hire of Jacob Trefethen, who led similar grantmaking at Coefficient Giving, a major funder of the effective altruism community that has periodically challenged OpenAI's AI vision. Wojciech Zaremba, an OpenAI co-founder, will serve as the foundation's head of AI resilience, focused on risks from more capable AI systems.
The timing of an expanded giving program, as OpenAI moves toward a potential IPO, is not lost on critics. OpenAI posted $13.1 billion in revenue in 2025 with a projected $8 billion (projected, not audited) operating loss for the year, per prior reporting. OpenAI President Greg Brockman has also co-funded a $125 million super PAC with Andreessen Horowitz and Palantir co-founders aimed at promoting AI-friendly policies, per Vox.
The $1 billion figure is real. What remains unsettled is whether a foundation structurally bound to the company it is meant to scrutinize can produce grantmaking that functions as genuine check on that company's direction. That is the question Brian Mittendorf, a professor of accounting and public affairs at The Ohio State University who studies nonprofits, said he watches most closely. People tend to focus on the financial piece, he told AP News. Is the immense value creation being used to further a charitable objective? But an equally important piece is whether the product they are developing is serving humanity as they envisioned.
The pledge will be managed by a new executive director the foundation is recruiting. Whether that person has genuine independence from the company the foundation governs is the first test of whether the $1 billion changes anything.