After two weeks of testimony painting him as a calculating opportunist, Sam Altman finally took the stand in the high-stakes legal battle between OpenAI and its co-founder, Elon Musk. For the jury, it was a study in contrasts: the image of a bewildered, “nice kid from St. Louis” facing off against a legal narrative that describes him as a “lying snake.”
Altman’s appearance was carefully calibrated. Entering the courtroom with a stack of evidence binders and a nervous energy that smoothed into a composed, credible delivery, he presented himself not as a titan of industry, but as a man perplexed by the accusations. When asked by his attorney, William Savitt, how it felt to be accused of “stealing” a charity, Altman’s response was pointed. “We created, through a ton of hard work, this extremely large charity, and I agree you can’t steal it,” Altman testified. “Mr. Musk did try to kill it, I guess. Twice.”
The testimony highlights a fundamental rift in the history of artificial general intelligence (AGI) development: the tension between a non-profit mission and the staggering capital requirements of modern compute. While Musk’s legal team has focused on the betrayal of OpenAI’s original charitable mandate, Altman’s defense leaned heavily on contemporaneous documents to argue that the shift toward a for-profit arm was a necessity—and one that Musk himself had attempted to steer toward his own total control.
The Battle for AGI Control
A central theme of Altman’s testimony was the fear of centralized power. Altman testified that during the early days of OpenAI, particularly following the company’s success with Dota 2, discussions regarding a for-profit entity became urgent. According to Altman, Musk insisted that any such entity be under his total control initially, asserting that only he could make the “non-obvious decisions” required for success.
Altman argued that this insistence on control contradicted the very reason OpenAI was founded: to ensure that no single individual could dominate the development of AGI. To illustrate this, Altman pointed to Musk’s own management style at SpaceX, citing a “hair-raising” conversation regarding succession. When asked who should lead OpenAI if he were to pass away, Altman testified that Musk suggested control should perhaps pass to his children.
This narrative of control is supported by a 2017 email from Altman to Shivon Zilis, in which he expressed concern that “no one person should have control of the world’s first AGI.” While Altman admitted to being open to “creative structures” to placate Musk in the short term, he maintained that Musk’s long-term goal was absolute authority—a pattern consistent with Musk’s history of avoiding situations where he could be ousted, similar to his departure from PayPal.
| Issue | Elon Musk’s Position | Sam Altman’s Testimony |
|---|---|---|
| Company Structure | OpenAI betrayed its non-profit mission. | For-profit arm was necessary for compute costs. |
| Governance | Altman “stole” the charity for profit. | Musk sought total personal control of AGI. |
| Tesla Integration | OpenAI should have merged with Tesla. | Merging would betray the mission; Tesla is a car company. |
| Funding | Non-profits can raise billions (e.g., Stanford). | Compute requirements exceed traditional philanthropy. |
The ‘Lightweight Threat’ and Financial Desperation
The trial also unearthed the strained dynamics between the founders as Musk began to distance himself from the organization. Altman revealed texts between himself and Sam Teller, in which Musk expressed a commitment to beefing up Tesla’s AI regardless of OpenAI’s trajectory. Altman described these messages as a “vague, like, a lightweight threat” that Musk would simply build the technology inside Tesla with or without the OpenAI team.
The financial reality of that period was stark. Once Musk ceased his quarterly donations, OpenAI was operating on what Altman described as a “shoestring” budget with an “extremely short runway of cash.” While the company sought other donors, the courtroom records revealed a complicated web of funding, including a donation from Alameda Research—the firm owned by Sam Bankman-Fried, who is currently imprisoned for fraud.
This financial precariousness, Altman argued, made the partnership with Microsoft not just an option, but a survival mechanism. He testified that throughout the process of establishing the for-profit arm, he kept Musk informed, and that Musk never formally objected to the moves privately, despite his later public grievances.
Legal Strategy: Reputation vs. Merit
The cross-examination of Altman, led by Steven Molo, shifted away from the legalities of non-profit law and toward a character assassination. For over ten minutes, Molo listed a litany of former colleagues and critics—including Ilya Sutskever, Mira Murati, and the founders of Anthropic—who had allegedly called Altman a liar. The strategy seemed designed to reinforce a public narrative of untrustworthiness rather than to dismantle the legal defense of OpenAI’s corporate restructuring.
One of the more contested arguments brought by Musk’s team was a comparison to Stanford University, suggesting that if a university can raise $3 billion a year as a non-profit, OpenAI should have been able to do the same. However, from a technical perspective, the comparison fails to account for the “compute bottleneck.” Unlike a university, which manages a donor network for scholarships and research, an AI lab requires billions in liquid capital specifically for massive GPU clusters to train frontier models. The initial Microsoft investments provided a scale of infrastructure that traditional philanthropy could not match.
Despite the aggressive cross-examination, Altman remained composed, appearing “hurt and confused” by the focus on his character. While this may have played well with the jury, the broader implications of the trial extend beyond the courtroom.
The legal battle has already begun to trigger secondary consequences. Reports indicate that several Republican Attorneys General and the House Oversight Committee are now looking into Sam Altman’s personal investments, with references to the trial’s revelations peppering those inquiries. Even if Altman wins the lawsuit, the trial has succeeded in cementing a public image of the OpenAI CEO as a polarizing figure in the race for AGI.
Disclaimer: This article discusses ongoing legal proceedings. The information provided is based on courtroom testimony and available evidence and does not constitute a legal judgment.
The court is expected to move toward closing arguments following the final submission of evidence binders. The next major checkpoint will be the jury’s deliberation on whether OpenAI breached its original founding agreement.
What do you think about the tension between non-profit missions and the cost of AI development? Share your thoughts in the comments or share this story on social media.
