On April 29, Elon Musk showed obvious impatience when he testified in court. The billionaire seemed annoyed when a lawyer for OpenAI asked him in court whether he had reneged on his commitment to invest in the startup in its early stages.
Musk has entered the second day of testimony in the high-profile trial. The case revolves around his accusation that OpenAI strayed from its altruistic mission in favor of profit. In court, the world's richest man fell into a passive defense position regarding his actual financial support for OpenAI since its inception.
Musk and OpenAI attorney William Savitt had many heated arguments over the rigor of the questioning, and he bluntly stated that the other party's questions were "unfair" and "misleading" several times.
Savitt repeatedly asked Musk what he called "simple" yes-or-no questions, but Musk said that such questions could not be answered so simply.
"Your question is not simple at all," Musk said in court. "Essentially, you are setting a trap to fool me."
Musk filed a lawsuit in 2024, accusing OpenAI CEO Sam Altman and President Greg Brockman of using tens of billions of dollars in Microsoft funds to transform the company into a for-profit enterprise and take the opportunity to enrich themselves.
OpenAI and Altman counter-sued Musk for malicious harassment, and bluntly stated that the real purpose of Musk's lawsuit was to suppress competitors and pave the way for xAI, an artificial intelligence startup he co-founded in 2023.
Savitt followed closely and asked Musk about the actual amount of investment in OpenAI during its initial stage. Musk revealed that he had been donating to OpenAI on a quarterly basis and paying rent for its office building until he later "lost confidence" in the company's management.
Musk co-founded the nonprofit with Altman, Brockman, and Ilya Sutskever, but the two parties have differed on the extent of Musk's financial support.
When OpenAI was announced in 2015, the nonprofit said Musk would eventually commit up to $1 billion to support its mission of developing artificial intelligence to "benefit humanity." Musk posted on social platform X in 2023 that he had actually donated US$100 million.
"At the end of the day, you didn't invest $1 billion in OpenAI, did you?" Savitt asked Musk.
Musk responded that around 2017, he had already had doubts about the development direction of OpenAI and gradually lost trust in the founding team.
Savitt immediately interrupted his speech. "My question is simple," he said, repeating the question.
U.S. District Judge Yvonne Gonzalez Rogers intervened in their dispute and asked Musk to answer Savitt's question head-on. Musk then asked the other party to repeat the question.
"My question is: Your actual investment in OpenAI is well under $1 billion, isn't it?" Savitt said.
"Strictly based on cash amount, I contributed a total of US$38 million." Musk replied.
The two had already fought in court. Savitt is one of the top corporate litigation lawyers in the United States. When Musk tried to launch a $440 billion acquisition of Twitter in 2022, it was he who sued on behalf of Twitter to require Musk to fulfill the acquisition agreement. In the end, Musk chose to compromise before the case went to trial.
The stakes in this Oakland court trial are high for OpenAI, and even related to the survival of the company. Musk has made multiple demands: claiming compensation of up to 134 billion US dollars, removing Altman and Brockman from their management positions, and canceling the profit-making transformation of OpenAI completed in October 2023.
Much of Musk's testimony so far has centered around his disagreements with OpenAI management. At the time, they were exploring strategies to raise enough money to compete with Alphabet's Google and other AI pioneers that operate on a for-profit model.
In 2017, Musk and other co-founders discussed setting up a for-profit subsidiary to provide financial support for artificial intelligence research and development. It was proposed that Musk hold a majority stake in the subsidiary and have almost absolute operational control.
According to the original plan, the board of directors has a total of 12 seats, of which Musk can appoint 4 seats, and Altman, Brockman, and Sutskwei each occupy 1 seat. Musk told the jury that the original plan was for his majority stake to be diluted quickly as more investors came on board.
But Musk admitted that the “last straw” that broke down the relationship was that the other co-founders seemed to oppose the proposal and expressed concerns about it, and the plan ultimately failed to materialize.
"They reneged on their previous agreement," Musk said. "I think this approach is very hypocritical. What they really want to do is to create a for-profit company and keep as much equity in their hands as possible."
Savitt showed the jury in court the records of email exchanges between Musk, Sutskwei, Brockman and others in 2017, which contained discussions surrounding the cooperation plan. In the email, Musk proposed that the development route needs to be adjusted to adapt to business needs.
In the same email exchange, Musk also offered to give Sutskevi and other OpenAI employees free Tesla cars.
Musk withdrew from the OpenAI board of directors in 2018 and founded xAI, a for-profit artificial intelligence company, five years later. Recently, xAI has been acquired by SpaceX, and the aerospace company is moving towards an initial public offering.
On the other hand, OpenAI’s valuation has approached one trillion U.S. dollars and it is also preparing to go public.
Musk told the jury that he established OpenAI as a nonprofit because he was concerned about the safety of artificial intelligence and wanted to ensure that the future of the technology was not completely controlled by companies like Google.
During the court hearing on April 29, Savitt asked Musk whether he believed that switching artificial intelligence companies to a profit-making model would create security risks. Musk said he believes it "does pose certain security risks."
"Then does the xAI you founded also have this kind of security risk?" Savitt asked.
"That's right." Musk replied.
Musk repeatedly emphasized throughout the testimony that he was not opposed to the establishment of for-profit artificial intelligence companies, but was firmly opposed to transforming a non-profit organization into a for-profit enterprise.
"It's like having your cake and eating it too, taking advantage of both sides," Musk said.
Savitt also raised a series of questions about Musk's conflict of interest between his role as a director of OpenAI and his involvement in Tesla and Neuralink, both of which have significant investments in AI.
The lawyer pointed out that although Musk had a fiduciary duty to OpenAI, in 2017 he secretly poached the top scientific research talents of the non-profit organization and tried to recruit them to other companies under his ownership.
Savitt said that Musk had recruited Andrej Karpathy, a well-known OpenAI scientist, to work at Tesla, and showed an email from Musk to Neuralink executives in court, which said, "You can go ahead and poach OpenAI talents to work at Neuralink. I have no objections."
"I think it's a free world," Musk responded, "and people should have the right to choose where they want to work."