NewsOpenAIMicrosoft

Nadella Takes the Stand in Musk-Altman Trial: 'We Have Each Other's Phone Numbers'

Microsoft's CEO testified that Musk never contacted him about concerns over OpenAI's direction. The testimony undercuts Musk's central claim in the case.

Alex Chen4 min read(Updated: )
Nadella Takes the Stand in Musk-Altman Trial: 'We Have Each Other's Phone Numbers'

Satya Nadella testified on Monday in the Musk v. Altman trial in San Francisco. His testimony, delivered over roughly three hours in federal court, undercut one of Elon Musk's core arguments: that Microsoft's multibillion-dollar investment in OpenAI was proof the company had abandoned its nonprofit mission, and that Musk was excluded from the process.

The trial background

Let me set the stage, because this case has a long and tangled history. OpenAI was founded in 2015 as a nonprofit with a mission to develop artificial general intelligence for the benefit of humanity. Elon Musk was a co-founder and early funder, contributing tens of millions of dollars. He left the board in 2018, citing a conflict of interest with Tesla's own AI ambitions. OpenAI then created a for-profit subsidiary, OpenAI LP, in 2019 to raise the massive amounts of capital needed to train increasingly large models. Microsoft invested $1 billion in 2019, and the relationship deepened from there.

Musk filed the lawsuit in early 2024, arguing that OpenAI had abandoned its nonprofit mission and that the Microsoft partnership amounted to a de facto acquisition without the formal transfer of ownership. His core claim: Sam Altman and the board enriched themselves through the for-profit conversion at the expense of the public-interest mission Musk says he funded.

The trial, unfolding in a San Francisco federal courtroom, has drawn intense attention because the stakes go far beyond two billionaires fighting. If Musk wins, it could force OpenAI to restructure or even unwind parts of its for-profit entity, with massive implications for the entire AI industry. Investors who poured billions into OpenAI at an $852 billion valuation would face serious questions about what they actually own. Other AI startups organized as for-profit subsidiaries of nonprofits, Anthropic among them, would face similar scrutiny.

What Nadella said

Nadella was direct. Musk never contacted him about OpenAI. Ever.

"We have each other's phone numbers," Nadella said from the stand. The implication landed hard. Musk, who had characterized Microsoft's involvement as a hostile takeover of a charity he co-founded, never picked up the phone to ask the CEO of Microsoft what was happening. If you were genuinely concerned that a massive corporation was corrupting a nonprofit you helped create, you would think a phone call might be step one.

Nadella characterized Microsoft's OpenAI investment, now totaling roughly $13 billion, as a "win-win." He said Microsoft shouldered "all the risk" when Musk stopped funding OpenAI and left the board in 2018. At that point, OpenAI was a small nonprofit with an ambitious research agenda and no clear path to sustainability. Microsoft provided the compute infrastructure and capital that turned GPT-3 from a research paper into a product.

He also addressed the governance question. Nadella acknowledged that Microsoft does not have a board seat at OpenAI, despite the size of its investment. "We structured this to maintain OpenAI's independence," he testified. "We wanted their mission to succeed, not to control it." This is a point Microsoft has made repeatedly, but hearing it under oath in federal court carries different weight than hearing it in a press release.

Why this testimony matters

The testimony matters because Musk's legal argument rests, in part, on the claim that OpenAI's partnership with Microsoft violated the original nonprofit charter. If Musk had concerns and never raised them with the counterparty who was allegedly enabling the violation, it makes the concern look less like a principled objection and more like litigation strategy.

There is also the credibility question. Musk has positioned himself as the defender of OpenAI's original mission, the founder who was pushed out and watched from the sidelines as the company he helped start was captured by corporate interests. Nadella's testimony paints a different picture: Musk left voluntarily, stopped funding the organization when it needed money most, and then, years later, sued when the company he left behind became enormously valuable without him.

The trial has become one of the most closely watched tech industry legal proceedings in years. Musk is seeking to unwind portions of OpenAI's for-profit conversion, arguing that Sam Altman and the board structured the Microsoft deal to enrich themselves at the expense of OpenAI's original mission. Altman's team argues Musk is a disgruntled former co-founder who left, founded a competitor (xAI), and now wants to slow OpenAI down through litigation.

Broader implications for AI industry governance

Beyond the courtroom drama, this trial is forcing a conversation the AI industry has been avoiding: what does nonprofit governance of for-profit AI companies actually mean, and does the structure have any teeth? OpenAI is not the only AI company organized this way. Anthropic is structured as a public benefit corporation with a long-term benefit trust that has the power to overrule the board on safety matters. Several other AI startups use similar hybrid structures.

If the court finds that OpenAI's for-profit conversion violated its nonprofit charter, every similar structure across the industry gets called into question. Investors will demand more clarity about what they actually own. Founders will face harder choices about how to structure their companies. And the public-interest mission language that has been standard in AI company founding documents since 2015 will face legal scrutiny it has never faced before.

Nadella's testimony did not close the case. There are weeks of trial left, and Musk's legal team has not yet presented its full case. But it did remove one narrative from the board. Musk was not shut out of conversations about OpenAI's future. He just never started them. Whether a jury finds that distinction legally meaningful remains to be seen, but it is a clean, factual point that Musk's team will have a hard time walking back.