OpenAI President Greg Brockman testified that Elon Musk demanded full control and CEO position in 2017, gifting co-founders Tesla Model 3s and reacting with anger when they refused, in a dramatic account of the founder's exit that TechCrunch reported and that exposes how OpenAI's early governance choices created legal vulnerabilities now playing out in Musk's lawsuit against the company.
Brockman's testimony came during the second day of Musk's trial in federal court in Oakland, where the Tesla and SpaceX founder is seeking to unwind OpenAI's for-profit structure and force it back to its nonprofit roots. Brockman, one of OpenAI's co-founders alongside Sam Altman, Ilya Sutskever, and Musk, recounted a 2017 meeting where Musk brought Tesla Model 3s as gifts for the other founders. Sutskever had commissioned a painting of a Tesla as a friendly gesture. The mood shifted when Musk was told the others would not give him a controlling stake or the CEO role. Brockman said Musk sat quietly for several minutes, then said, "I decline," stood up, stormed around the table, grabbed the painting, and stormed toward the door. He turned and asked, "When will you be departing OpenAI?" Brockman thought Musk was going to hit him.
The account fills in details of a pivotal moment that has shaped the AI industry. Musk co-founded OpenAI in 2015 as a nonprofit lab to develop artificial general intelligence for humanity's benefit. By 2017, disagreements over funding, control, and commercialisation had surfaced. Musk wanted 51 percent equity and the CEO role, arguing he deserved it because he had started multiple multi-billion-dollar companies. Brockman said Musk told them, "You guys are great, but I can start another AI company tomorrow, like in one tweet." The other founders declined, and Musk stopped regular donations, left the board in February 2018, and later founded xAI. Brockman testified that OpenAI had over $150 billion in equity value, built through hard work since Musk left.
Musk's lawyers have focused on journal entries from Brockman and Sutskever discussing whether to remove Musk from the board, including a line about it being "morally bankrupt" to "steal the nonprofit" from him. Brockman said those entries were about exploring options but they ultimately did not act. Musk's suit alleges Altman and Brockman betrayed the founding agreement by prioritising Microsoft over the public good and turning OpenAI into a closed for-profit. Brockman countered that Musk knew about plans to create a for-profit subsidiary to raise capital needed to compete with Google.
OpenAI's hybrid structure, with a nonprofit controlling a capped-profit subsidiary, was meant to balance mission and incentives. Brockman's account shows how that balance broke down early. Musk wanted absolute control to steer the lab's direction. The others wanted distributed governance to prevent any single person from dominating. The nonprofit board later fired Altman in November 2023 citing lack of candor, only to reverse under employee and Microsoft pressure. That episode echoed the 2017 tensions, with personal trust eroding as commercial stakes rose. The governance model worked when OpenAI was a small research lab but struggled as it became a capital-intensive platform requiring billions in compute and revenue in the tens of billions.
For San Francisco founders, the dispute teaches that mission-driven companies face acute governance risks when they later need enormous compute capital. OpenAI's founding documents promised open research and broad benefit with capped investor returns. Scaling to frontier AI required rewriting those terms, raising billions from Microsoft, and prioritising commercial deployment over pure research. The nonprofit board lacked the expertise to manage executive performance or strategic pivots. The for-profit arm became dominant, leaving the nonprofit as symbolic oversight. Brockman's testimony shows those structural misalignments created day-to-day chaos, with founders managing by division rather than consensus.
The compute ambition is the underlying pressure. Brockman testified OpenAI plans to spend $50 billion on computing power in 2026 alone. That scale requires for-profit capital markets, not nonprofit donations. The mission promise becomes harder to honour as the company needs the same commercial infrastructure as any trillion-dollar tech firm. Founders promising public benefit or open research should design boards, incentives, and decision processes that can withstand that transition. OpenAI did not, and the legal cost is now material. Musk's suit may not succeed, but it has already forced OpenAI to litigate its own founding story in public.
Also read: Anthropic's dreaming agents introduce governed self-improvement as the next enterprise battleground • Genesis AI's full-stack robotics bet says the moat is not the model, it's owning everything around it • Mira Murati's court testimony that Altman misled her exposes OpenAI's governance under extreme pressure
