Even as Elon Musk and Sam Altman battle it out in court over OpenAI’s transformation to a for-profit entity, some previously unknown facets of the company’s early days are also coming to light.
Testimony from Greg Brockman, OpenAI’s co-founder and former President, has revealed that in 2017, Sam Altman personally compensated him with approximately $10 million in equity — drawn not from OpenAI’s treasury, but from Altman’s own private family office. Brockman confirmed the arrangement under oath, stating that Elon Musk, OpenAI’s other prominent co-founder and primary early backer, was never informed.
The Deal That Stayed Hidden
The arrangement came to light only after Jared Birchall, Musk’s head of family office, discovered it and wrote to Musk in August 2017. In his email, Birchall described the compensation as a stake in the LLC holding the assets of Altman’s personal family office, worth $10 million. More pointedly, he flagged what he saw as the logical consequence: that Brockman would naturally have greater allegiance to Altman as a result.
Musk’s response was characteristically terse — he forwarded the email to Brockman with just two question marks. Brockman, according to court proceedings, did not respond.
When asked directly under oath whether Musk knew about the arrangement, Brockman said: “He certainly didn’t know about it.” When pressed on whether he believed it was fair to have a side deal concealed from a co-founder, Brockman fell back on a procedural defense: “We had never discussed compensation directly.” His explanation for not disclosing it to Musk? That Musk’s time was “relatively hard to get.”
Why This Is Unusual
In the world of startups and co-founded ventures, compensation arrangements between co-founders are almost always structured transparently — especially when those arrangements could influence loyalty and decision-making at the leadership level. It is highly unusual, and widely considered a breach of co-founder trust, for one founding partner to privately compensate another without the knowledge of remaining co-founders.
The reason is straightforward: co-founders rely on each other to act in the collective interest of the organization. Hidden financial ties between two of them — particularly when one is being compensated by the other — create an undisclosed conflict of interest. In OpenAI’s case, this was a nonprofit at the time, making the off-books nature of the deal even more striking.
The Allegiance Question
The Birchall email cuts to the heart of the governance concern: if Brockman held $10 million in equity tied to Altman’s personal wealth, his incentives were no longer purely aligned with the organization or its co-founders collectively — they were partially tethered to Altman personally.
Brockman’s own response on the stand was telling. Asked whether he was loyal to Altman, he said, “I don’t know I would say it quite like that.” It was a careful non-denial — neither confirming nor deflecting the underlying implication. The arrangement, as structured, meant that Brockman’s financial fortunes were in part linked to Altman’s, not to Musk’s, and not to OpenAI’s stated charitable mission.
For Musk — who was a primary funder of OpenAI’s early operations and had brought significant credibility and capital to the venture — the existence of such an arrangement without his knowledge raises a legitimate fairness question. A co-founder of Musk’s stature, putting real money into a nonprofit AI lab, had a reasonable expectation that the leadership structure around him was free of undisclosed financial entanglements.
Context Within the Broader Trial
The revelation lands in the middle of Musk’s lawsuit against OpenAI, which centers on his claim that the organization abandoned its founding nonprofit mission in favor of commercial interests. Musk argues he was misled about the direction OpenAI would take.
The Brockman compensation arrangement, whatever its original intent, adds texture to that narrative. It suggests that even in 2017 — years before OpenAI’s for-profit pivot became public — the dynamics within OpenAI’s leadership were less transparent than they appeared. Altman was building personal financial ties with at least one co-founder while Musk remained unaware.
Whether or not this arrangement influenced any specific decisions at OpenAI, courts and observers will have to weigh what it reveals about the culture of trust — or the absence of it — at the organization’s founding table.