Court testimony and internal emails presented during the Musk v. Altman trial reveal that Elon Musk actively recruited OpenAI researcher Andrej Karpathy to Tesla's self-driving AI team in late 2017, directly contradicting Musk's earlier sworn statement that Karpathy had already planned to leave. The evidence suggests Musk was systematically attempting to poach top talent from OpenAI while still serving on its board, raising questions about his true intentions for the nonprofit organization he co-founded.

What Did the Trial Evidence Show About Karpathy's Recruitment?

During cross-examination of Shivon Zilis, a former OpenAI adviser and board member, OpenAI's legal team presented text messages and emails from June 2017 showing that Musk had directly approached Karpathy about joining Tesla. When Zilis was informed that Karpathy had signed an official offer, she responded with enthusiasm, texting "Fuck yeahhhhhhh" to Tesla colleagues. When asked whether OpenAI would be upset about the hire, Zilis acknowledged it was a fair question but claimed she had "talked to Greg [Brockman] today and he clearly had no idea".

Zilis

The significance of this evidence became clear when OpenAI's lawyer William Savitt pointed out that Zilis' testimony directly contradicted what Musk had told the jury just days earlier. Musk had testified that Karpathy left OpenAI of his own volition, but the documentary evidence suggested otherwise. In a 2017 email to a Tesla vice president, Musk himself wrote: "The OpenAI guys are gonna want to kill me. But it had to be done". This statement implies Musk knew the recruitment would be viewed as a betrayal by OpenAI leadership.

How Did Karpathy Fit Into Musk's Larger Tesla AI Strategy?

Karpathy's recruitment was part of a much broader effort by Musk to build what he called a "world-class AI lab" at Tesla that would rival Google DeepMind and Facebook AI Research. In November 2017, Zilis drafted a frequently asked questions document for a Tesla event at the NeurIPS AI conference that listed Karpathy as a planned leader of the unit, alongside Musk himself. The document explicitly stated that "one major issue for Tesla is when people think of Elon and AI, they think of OpenAI".

Notably, the FAQ also included Sam Altman's name next to Musk's with two question marks, suggesting Musk was simultaneously trying to recruit OpenAI's CEO to join Tesla's AI effort. One annotation in the document noted that Altman could be a moderator for the NeurIPS event, which "could be a forcing function for Sam to commit to TeslaAI". Neither Altman nor the broader Tesla AI lab initiative materialized as planned, but the evidence reveals the scope of Musk's ambitions to consolidate AI talent under Tesla's control.

What Pattern Does This Reveal About Musk's OpenAI Strategy?

The Karpathy recruitment is one piece of a documented pattern showing that Musk's involvement with OpenAI was driven by a desire for control rather than a commitment to the nonprofit's stated mission of ensuring artificial general intelligence benefits all of humanity. According to trial testimony from OpenAI president Greg Brockman, Musk began pushing for OpenAI to convert to a for-profit structure in mid-2017, but with a critical condition: he wanted full control. When that equity discussion didn't go his way during an in-person meeting, Brockman testified that Musk stood up, stormed around the table so aggressively that Brockman feared he would be physically attacked, tore a painting off the wall, and threatened to withhold funding.

Zilis' emails and testimony revealed additional attempts by Musk to either absorb OpenAI into Tesla or gain control of its direction. In one email, she wrote that OpenAI's cofounders had not "internalized the advantages of burying this in Tesla for stealth advantage". Another email from February 2018 proposed several scenarios for creating a counterbalance to Google DeepMind, including having Altman run a Tesla AI lab and even attempting to recruit Demis Hassabis, the leader of Google DeepMind, to Tesla. Zilis wrote: "Find a way to get Demis. Seriously... If he hung around E perhaps it would force him to think about humanity more".

Zilis

Steps to Understanding the Control Pattern in Musk's AI Ventures

• The Funding Leverage: Musk used his position as a major donor and board member to pressure OpenAI toward decisions that would benefit him, including halting his $5 million quarterly funding commitment in August 2017 without immediately informing the organization, according to Zilis' emails.
• The Talent Acquisition: Rather than building AI capabilities at Tesla independently, Musk systematically recruited top researchers from OpenAI, including Karpathy, while still serving on OpenAI's board, creating a conflict of interest that he later framed as a reason to resign.
• The Narrative Control: When Musk left OpenAI's board in February 2018, he publicly claimed it was due to a conflict of interest with Tesla's AI work, but trial evidence suggests he departed because he failed to secure the CEO role and couldn't achieve the level of control he demanded.

The trial has exposed what many observers suspected: Musk's involvement with OpenAI was never primarily about advancing AI safety or ensuring the technology benefited humanity. Instead, it was about positioning himself to control one of the most important technological developments of the era. When he couldn't achieve that control through OpenAI, he left and eventually created xAI, a private company under his complete ownership.

"There was documentary evidence that, at several points, Mr. Musk had contemplated seeking to join Sam Altman to the board and offered that option. It was part of Mr. Musk's effort to corrupt OpenAI and absorb it into Tesla. He was trying to get Altman to abandon the mission and be part of Tesla," said William Savitt, OpenAI's lawyer.

William Savitt, OpenAI Attorney

The Karpathy recruitment case illustrates a fundamental tension in Musk's approach to AI development. While he has publicly positioned himself as concerned about AI safety and the need for open access to advanced AI systems, his actions in 2017 and 2018 suggest his primary motivation was securing competitive advantage and personal control. The trial evidence indicates that when OpenAI's cofounders refused to give him the control he demanded, Musk simply walked away and built his own AI company, taking talent and resources with him.

As the trial continues, the pattern of Musk's behavior toward OpenAI raises broader questions about how AI development should be governed and whether individuals with outsized influence over multiple AI initiatives should be allowed to maintain conflicting interests. The Karpathy recruitment, once a footnote in AI industry history, has become central evidence in a legal battle that may reshape how the industry thinks about conflicts of interest, nonprofit governance, and the concentration of AI talent and resources.