Elon Musk returned to the witness stand on Wednesday, continuing his testimony in the high-stakes legal battle against OpenAI and its CEO, Sam Altman. Under rigorous cross-examination by OpenAI’s legal team, Musk was pressed on a series of actions he allegedly took to exert control and disrupt the organization during a pivotal power struggle in 2017. Evidence presented, including emails, suggests Musk attempted to poach OpenAI researchers and withheld promised funding during this period, ultimately losing the internal dispute over the company’s direction. The proceedings have laid bare the early tensions and strategic maneuvers that shaped one of the world’s most influential artificial intelligence organizations.
The courtroom atmosphere was palpable with tension as the cross-examination commenced. Judge Yvonne Gonzalez Rogers began the day by issuing a reprimand to an individual in the gallery for attempting to photograph Musk. Greg Brockman, OpenAI’s president and co-founder, observed Musk’s testimony from behind his legal counsel, his gaze fixed on the witness with a noticeable intensity. Musk himself appeared increasingly frustrated, frequently pausing to challenge the questions posed by OpenAI’s attorney, William Savitt, whom he accused of presenting misleading inquiries. Savitt’s line of questioning, however, was repeatedly punctuated by objections from Musk’s counsel, technical glitches, and Musk’s consistent claims of an inability to recall specific details regarding OpenAI’s early history.
The 2017 Power Struggle: Emails Reveal Musk’s Bid for Control
Central to the legal proceedings are emails exchanged in September 2017 between Musk, Brockman, and researcher Ilya Sutskever. These communications, presented as evidence, detail discussions surrounding the formation of what would eventually become OpenAI’s for-profit arm. In a particularly contentious exchange, Musk asserted his demand for the right to appoint four members to the board of directors, a move that would have granted him greater voting power than his co-founders, who would collectively hold three seats. "I would unequivocally have initial control of the company, but this will change quickly," Musk stated in one of the messages. Sutskever, in his response, rejected the proposal, citing concerns that it would grant Musk excessive authority. This demand for ultimate control foreshadowed a broader ideological rift that would soon divide the founding members.
Funding Withheld: Musk’s Financial Pressure Tactics
The timing of these negotiations is significant, as it coincided with Musk’s decision to halt payments to OpenAI. This cessation of funding proved particularly challenging for the organization, given Musk’s position as its primary financial backer at the time. Since OpenAI’s inception in 2016, Musk had committed to quarterly payments of $5 million as part of a broader $1 billion pledge. However, in the spring of 2017, these disbursements ceased. Further illuminating this strategy, an email from August 2017 shows Jared Birchall, the head of Musk’s family office, seeking Musk’s confirmation on whether to continue withholding the funds. Musk’s terse reply was a definitive “Yes.” This period marked a critical juncture, demonstrating Musk’s willingness to leverage financial leverage in his pursuit of organizational control.
Poaching Talent: Musk’s Alleged Recruitment Efforts
In parallel with the power struggle, emails indicate that Musk engaged in discussions with executives at Tesla and Neuralink, his brain-computer interface company, regarding the recruitment of OpenAI employees. Notably, at this time, Musk remained a member of OpenAI’s board of directors.
One email, sent in June 2017 to a Tesla vice president, detailed Musk’s outreach to Andrej Karpathy, an early OpenAI researcher. "Just talked to Andrej and he accepted as joining as director of Tesla Vision," Musk wrote. He further elaborated on Karpathy’s expertise: "Andrej is arguably the #2 guy in the world in computer vision… The openai guys are gonna want to kill me, but it had to be done." On the witness stand, Musk attempted to mitigate the implications of this action, asserting that Karpathy was already contemplating leaving OpenAI. "Andrej had made his decision. If he’s going to leave OpenAI, he might as well work at Tesla," Musk stated.
Additional evidence surfaced in October 2017, when Musk communicated with Ben Rapoport, a co-founder of Neuralink. Musk instructed him to "Hire independently or directly from OpenAI. I have no problem if you pitch people at OpenAI to work at Neuralink." When confronted with these communications, Musk argued that restricting employment opportunities would have been illegal. "It’s illegal to restrict employment. It would be illegal to say you can’t employ people from OpenAI. You can’t have some cabal that stops people from working at the company they want to work at," he contended.
Shivon Zilis and the Shifting Dynamics
Further complicating the narrative, a text message from February 2018 between Musk and Shivon Zilis, a then-OpenAI board member and an executive at Neuralink, was presented. Zilis is also the mother of four of Musk’s children. In the exchange, Musk stated, "We are going to actively try to move three or four people from OpenAI to Tesla. More than that will join over time, but we won’t actively recruit them." When Zilis inquired about maintaining her relationships within OpenAI while still serving on the board, Musk advised her to remain "close and friendly." This communication suggests a strategic approach to personnel movements and ongoing engagement with OpenAI, even as Musk pursued talent for his other ventures.
Broader Context and Implications
The legal battle stems from Musk’s lawsuit alleging that OpenAI has strayed from its original mission of benefiting humanity and has instead become a de facto subsidiary of Microsoft. Musk argues that the company’s pursuit of profit and its partnership with Microsoft violate the founding principles of OpenAI, which was initially established as a non-profit research organization. His legal team seeks to compel OpenAI to revert to its non-profit status and to release its research for the public good.
OpenAI, in its defense, contends that Musk was aware of and supported the organization’s shift towards a for-profit structure. They argue that his current lawsuit is a retaliatory measure stemming from his loss of influence and control within the company. The evidence presented in court, particularly the emails detailing Musk’s demands for power and his subsequent actions regarding funding and personnel, forms a critical part of OpenAI’s defense.
The case has significant implications for the future of AI development and governance. It raises fundamental questions about the balance between commercial interests and the ethical responsibilities of powerful AI organizations. The outcome could set precedents for how such companies are regulated and how their missions are interpreted and enforced. Furthermore, the detailed examination of early internal communications provides a rare glimpse into the formative stages of a technology that is rapidly reshaping the global landscape.
Musk’s cross-examination is scheduled to continue, and further testimony is expected from key figures including Jared Birchall and Greg Brockman. The proceedings are anticipated to delve deeper into the intricate relationships and strategic decisions that have defined the trajectory of OpenAI and its founder’s evolving role within the AI ecosystem. The court’s eventual ruling will not only resolve this specific dispute but may also cast a long shadow over the broader development and deployment of artificial intelligence.
Correction: April 29, 2026, 7:00 PM PDT: The dates of several emails presented to the court have been corrected.
This article is an edition of Maxwell Zeff’s Model Behavior newsletter. Read previous newsletters here.
