The high-stakes legal battle between Elon Musk and OpenAI, particularly CEO Sam Altman, is currently unfolding in federal court, revealing a dramatic saga of ambition, competing visions, and alleged betrayals. At the heart of the *Musk v. Altman* trial lies Musk’s assertion that Altman and OpenAI President Greg Brockman effectively hijacked a nonprofit organization, using his initial $38 million investment to transform it into a private entity now valued at over $800 billion. However, recent courtroom revelations suggest a complex backstory, with OpenAI’s legal team questioning Musk’s own past attempts to control the very organization he now sues.
New evidence, including emails and compelling testimony, paints a picture of Musk’s intense efforts to integrate OpenAI’s leadership into his own ventures, specifically Tesla. These revelations come amidst concerns raised by Musk’s lawyers regarding Altman’s alleged history of deceit, citing video depositions from former OpenAI CTO Mira Murati and board member Helen Toner.
Elon Musk’s Vision for a Tesla AI Powerhouse
Months before his departure from OpenAI’s board in February 2018, Elon Musk made a significant play to recruit Sam Altman. He allegedly offered the then-OpenAI CEO a coveted seat on Tesla’s board, envisioning Altman leading a “world-class AI lab” within Tesla itself. This astonishing detail emerged during the cross-examination of Shivon Zilis, a former OpenAI adviser, Neuralink and Tesla executive, and mother to four of Musk’s children.
The court saw a text from Zilis to Altman in February 2018, asking, “Did you think through a B Corp subsidiary of Tesla?” This suggests a serious consideration of how OpenAI’s mission might be absorbed or replicated under Tesla’s umbrella. OpenAI’s lawyer, William Savitt, described this as part of “Mr. Musk’s effort to corrupt OpenAI and absorb it into Tesla,” aiming to sway Altman from the nonprofit’s original mission.
Further evidence showcased Tesla’s ambitious plans, including a drafted FAQ for a November 2017 event at the NeurIPS AI conference. The document openly stated its purpose: “to share that Tesla is building a world leading AI lab(?) which will rival the likes of Google / DeepMind and Facebook AI Research.” A key concern noted in the draft was that “when people think of Elon and AI, they think of OpenAI,” highlighting Musk’s desire to shift that perception towards Tesla.
The drafted FAQ even listed potential leaders for this new Tesla AI unit, including Musk himself, former OpenAI researcher Andrej Karpathy, and notably, “Altman??” with two question marks. A handwritten note suggested that having Altman moderate the NeurIPS event “could be a forcing function for Sam to commit to TeslaAI.” While Zilis testified that Altman never joined Tesla and the AI lab never materialized, these documents reveal the depth of Musk’s prior intentions.
The Battle for AI Supremacy: Absorption or Rivalry?
The trial also delved into Musk’s efforts to recruit other key talent, including Andrej Karpathy, who later joined Tesla. In June 2017, Zilis reacted with “Fuck yeahhhhhhh” upon learning Karpathy had signed with Tesla. When asked if OpenAI would be upset, Zilis noted she had spoken to Greg Brockman, who “clearly had no idea,” suggesting a deliberate strategy to recruit from OpenAI without their knowledge. This directly contradicts Musk’s earlier testimony that Karpathy left OpenAI of his own volition.
Another fascinating aspect of the trial involves the concept of “burying” OpenAI within Tesla. In an October 2017 email, Zilis noted that OpenAI’s co-founders had not “internalized the advantages of burying this in Tesla for stealth advantage.” While Zilis denied this meant closed-sourcing Tesla’s AI, explaining it as a “small fish in a big pond” scenario, it underscores Musk’s strategic thinking about how to position his AI efforts against competitors.
Musk, viewing Google DeepMind as a major threat, explored various scenarios to create a formidable counterbalance. In a February 2018 email to Musk, Zilis outlined these possibilities:
- Having Sam Altman run a Tesla AI lab.
- Actively pursuing Demis Hassabis, DeepMind’s leader, to “come to Tesla somehow or DeepMind is spun out.”
These discussions reveal Musk’s consistent drive to either acquire top AI talent or absorb leading AI initiatives into his own ecosystem, long before he launched his rival AI lab, xAI.
Shivon Zilis: A Board Member’s Complicated Position
Shivon Zilis’s testimony has proven particularly pivotal, not just for her insights into Musk’s past plans, but also due to her unique personal connection. She was appointed to OpenAI’s board of directors in January 2020. During her tenure, Zilis became pregnant with Musk’s children via IVF but did not disclose this to other board members due to a confidentiality agreement with Musk.
The conflict of interest eventually came to light when Business Insider contacted her in 2022 about the children. Zilis testified that she subsequently informed Altman, but her board service continued until February 2023. Her resignation coincided with Musk’s public announcement of xAI, his competing AI venture. While Zilis claimed she learned about xAI from Altman, text messages presented in court suggest she was already aware of Musk’s plans.
In a text to a friend on the day of Altman’s call, Zilis wrote, “Have to resign [from the] openai board btw. [Musk’s] effort has become well known.” She added, “When the father of your babies starts a competitive effort and will recruit out of openai there is nothing to be done.” Her testimony has provided a rare glimpse into the intricate web of relationships and high-stakes maneuvering at the forefront of the AI industry.
The trial continues with anticipation, as lawyers for Musk are expected to call additional witnesses, including Rosie Campbell, a former OpenAI employee, and David Schizer, a specialist in nonprofit law. This ongoing legal saga not only illuminates the origins and evolution of OpenAI but also the intense competitive landscape defining the future of artificial intelligence.
Source: Wired – AI
