MONDAY MARKED THE BEGINNING of the third week of trial in Elon Musk’s lawsuit against OpenAI and its leaders, Sam Altman and Greg Brockman, and their co-defendant, Microsoft Corporation.
Trial testimony on Monday was presented by two witnesses — Microsoft CEO Satya Nadella and OpenAI former chief scientist Ilya Sutskever — who have been central in the testimony of earlier witnesses.
The testimony was part of a suit brought forward by Musk which contends that OpenAI was formed as a nonprofit with a mission to develop artificial intelligence for the benefit of humanity, but that over time it abandoned its mission in favor of enriching the insiders and its “strategic partner” Microsoft.
Nadella’s testimony
Prior to Nadella’s testimony, Microsoft has tried to lie low during the trial, its lawyers popping up from time to time to remind the jury that much of the fight between Musk and Altman came before it arrived on the scene and that Microsoft had nothing to do with it.
Nadella testified for two and a half hours, a good part of it about a dramatic weekend in November 2023, when Altman and Brockman were fired by OpenAI’s nonprofit board, only to be reinstated a few days later after Microsoft announced that it was offering employment to all OpenAI’s employees.
Nadella stood tall and trim in a black suit as he took the oath. On his direct examination, he delivered his testimony in a well-polished manner. He described the beginning of his company’s relationship with OpenAI as a loss-leader arrangement in which Microsoft offered discounted services with the hope of a marketing benefit.
However, after OpenAI created its for-profit subsidiary, Microsoft made three investments in OpenAI — one in 2019, one in 2021 and one in 2023. Each of the transactions involved a technology component and an investment component. Nadella characterized the transactions as steps in a deepening relationship that evolved over time.
Nadella described what each party contributed and received across the three deals.
Microsoft provided OpenAI with $13 billion in investment, custom-built supercomputers and cloud computing, and it also allowed OpenAI to retain the rights to commercialize its own technology despite Microsoft’s exclusive license. Those were things integral to the nonprofit’s mission, as Nadella understood it.
Microsoft received OpenAI “know-how,” an “exclusive” license to use and make OpenAI’s technology commercially available and a financial return from three different revenue streams — payments for using Azure cloud services, a profit share and a return on its equity investment.
Nadella said Microsoft was a for-profit company and that its responsibility was to its shareholders. The company negotiated for the best deal it could get, he said, and that it took a big risk, particularly in making the largest of its three investments, which was $10 billion in 2023. While he believed the investment was prudent, there was a risk it could have been lost, he said, further noting that he wanted to protect the investment and fight for it.
Musk’s team used its cross examination of Nadella to show that Microsoft had progressively obtained greater and greater access and control of the technology and the value it created.
Nadella said that, for Microsoft, getting access to OpenAI’s technology was central to its willingness to invest. Through the three transactions, Microsoft got an “exclusive” license to develop commercial products from OpenAI’s technology (though OpenAI also retained the right to do so). It obtained the right to embed 20 full-time Microsoft employees at OpenAI. In the future, Microsoft is entitled to have an embedded employee for every 10 OpenAI employees.
Microsoft also obtained the right to participate in “major decisions” by OpenAI that affected its commercialization of the technology.
In their initial agreements with OpenAI, Microsoft did not get any interest in OpenAI technology after AGI (or artificial general intelligence) was reached. However, a corporate restructuring in 2025 added those rights. At the same time, the cap on its returns from its investment was eliminated. Nadella admitted that today Microsoft holds more of a stake in the for-profit than the nonprofit does but said it was because his company took a risk.
He added that he was “proud Microsoft took the risk to sponsor a fledgling company so that its technology could make a difference.”
Nadella smiled as he walked from the courtroom, giving Brockman, OpenAI’s president, a friendly tap on the shoulder as he passed by.
Sutskever: OpenAI founder, former chief scientist
Also taking the stand for the first time Monday was Ilya Sutskever, former chief scientist at OpenAI. Sutskever was a member of OpenAI’s founding team, recruited to nonprofit OpenAI from Google in late 2015 after what he called “a big team effort” by Musk, Brockman, Altman and others.
Sutskever said that at the time, he viewed Musk as one of the most capable CEOs in the world.
In August 2017, OpenAI’s artificial intelligence bot had beaten an elite human player in an international video game competition. It was, he said, “a big public accomplishment” conveying that OpenAI could compete with Google’s DeepMind artificial intelligence lab.
Sutskever testified that OpenAI’s founders discussed ‘outlandish ideas’ to acquire more computing power, including for-profit alternatives. Musk didn’t object, but insisted on control and a majority of the economic interests.
The victory led to the “realization” that to make further progress, OpenAI would “need a big computer,” a computer comparable to the human brain with its billions of neurons and synapses. The founders — Musk, Altman, Brockman and Sutskever — began to discuss “lots of outlandish ideas” to acquire that computing power, including for-profit alternatives. He testified that Musk did not object to the prospect of a for-profit structure but insisted on control and a majority of the economic interests.
Ultimately, when the founders were unable to agree on a new structure, Musk left the OpenAI board in early 2018. The others went forward with the nonprofit.
In 2019, the nonprofit created the for-profit subsidiary that inked the Microsoft contracts Nadella described.
There was rapid change over that period in tech capabilities. The difference in OpenAI’s models between 2017 and 2024 was, Sutskever said, “like the difference between an ant and a cat.”
Sutskever testified about his role in the events of November 2023 when Altman was fired. Sutskever was a member of OpenAI’s nonprofit board of directors at that point and prepared a report at the request of other board members documenting serious problems with Altman’s leadership as CEO.
The report accused Altman of “a consistent pattern of lying” and the “creation of an environment not conducive to the development of safe artificial general intelligence.” The board relied on the report in part when its members voted 4-0 to fire Altman.
Five days later, after the chaotic weekend in which hundreds of OpenAI employees said they would leave to go Microsoft, the board, including Sutskever, voted to reinstate Altman as CEO. When asked why he’d reversed himself, Sutskever said he was concerned that OpenAI would be destroyed by the instability that was caused by the firing.
“It was a Hail Mary,” he said.
Sutskever left OpenAI in May 2024. Before he left, OpenAI had formed a “Superalignment Team” to research the technical means to make “unbelievably powerful AI” controlled and safe. He said the team was doing “the most important work for the long term.” The Superalignment Team was disbanded shortly after Sutskever left the organization.
Tuesday and upcoming days
Brett Taylor, chair of the OpenAI Foundation, the name of the nonprofit OpenAI adopted after 2025, testified after Sutskever but did not complete his direct testimony and will continue on Tuesday. Altman is also expected to testify on Tuesday. The judge expects all trial testimony to be finished on Wednesday with closing arguments on Thursday. There will be no trial on Friday, and jury deliberations will commence on Monday.