During the Musk v. Altman trial, Greg Brockman testified that his OpenAI stake is worth approximately $30 billion. It was worth considerably less when he co-founded the company in 2015 — he joined as the third employee, essentially employee #3. So a Musk attorney asked the obvious question: why hasn't he donated $29 billion to OpenAI's nonprofit arm?
The Structure Was Designed This Way
OpenAI's original structure was explicit: a nonprofit controlling a capped-profit entity, designed to ensure the technology would benefit humanity rather than enrich investors. But that structure collapsed when Microsoft invested billions and the company pivoted to a "capped-profit" that now looks a lot like a regular startup — except for the $30B in paper value sitting on co-founder balancesheets.
The uncomfortable truth is that the nonprofit was always aPR move. It attracted talent, deflected regulatory scrutiny, and gave the company moral high ground. But when the money materialized, the structure bent rather than broke. OpenAI is now worth $10B pre-money in a new joint venture, and Brockman is sitting on a stake almost as large as Stripe's valuation when it went public.
The nonprofit was always a PR move. It attracted talent, deflected regulatory scrutiny, and gave the company moral high ground. But when the money materialized, the structure bent.
Why This Matters Beyond the Trial
Every AI company claiming to be "safety-first" or "for humanity" should answer the same question. The structural distance between stated mission and actual incentives is the gap where regulatory backlash gets bred. When founders hold billions but the nonprofit holds empty pockets, regulators notice.
Already, the Trump administration is discussing an executive order to create an AI working group that vets models before release — a direct response to exactly this kind of concentration. Former AI advisers from both parties are urging tighter export controls and safety audits. The message is clear: if you won't self-regulate, government will.
The $29B question isn't really about Greg Brockman. It's about whether any AI company can claim public benefit while the founders capture private upside. The trial ends, but the question stays open.