Elon Musk attended a federal court in California on Wednesday to claim that Sam Altman and his cofounders “stole a charity.” In an unexpected twist, he admitted that Tesla is not currently pursuing artificial general intelligence (AGI), contradicting a recent tweet.
Musk’s lawsuit against OpenAI’s structure alleges that Altman and others misled him about the nonprofit’s intentions and then allowed the for-profit branch to dominate. Musk’s hours-long testimony revealed that the case might hinge on the importance of profit caps for OpenAI investors. Musk co-founded OpenAI with Altman and others, initially trusting them to create AI for humanity. Over time, he grew wary of their intentions and accused them of exploiting the nonprofit.
OpenAI lawyer William Savitt challenged Musk’s account, highlighting his role in OpenAI’s shift to for-profit status to secure necessary investments, including a potential incorporation with Tesla. Musk considered turning OpenAI for-profit as early as 2016 and explored owning majority equity in a for-profit arm in 2017. These plans failing led to Musk ceasing regular donations while continuing to cover office expenses until 2020.
Musk argued that there is a significant difference between capped and uncapped investor profits. Microsoft’s early investments in OpenAI had profit limits, which have relaxed over time, prompting Musk’s lawsuit.
Savitt attempted to demonstrate Musk’s involvement in later fundraising ventures and his lack of objection, noting that one of his advisers, Shivon Zillis, was also on OpenAI’s board.
In cross-examination, Musk’s statement that Tesla’s AI focuses solely on self-driving, not AGI, was confronted with a tweet claiming a future AGI pursuit. Musk clarified in court, “We are not pursuing AGI right now,” which differs from his public post.
Musk’s claim of a $100 million investment in OpenAI, exceeding the actual $38 million, was discussed, with Musk arguing his influence compensated for the difference. Savitt highlighted Musk’s support for Tesla and Neuralink’s recruitment of OpenAI staff, including his efforts to hire OpenAI leaders after leaving the board in 2018.
The discussion shifted to AI safety concerns, central to Musk’s argument that OpenAI’s corporate transition poses risks. Savitt had Musk concede that AI companies, including his own, face such risks. Judge Gonzalez Rogers allowed this topic to continue with some restrictions and dismissed unrelated scandals.
Musk’s next court appearance will feature more questioning. Testimonies are also expected from Musk’s family office manager Jared Birchall, AI safety expert Stuart Russell, and OpenAI president Greg Brockman.
Correction: A previous version of this story contained errors regarding the Tumbler Ridge shooting and has been updated.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.