While Musk was away on a trans-Pacific business trip, Altman, who made his first court appearance in the "OpenAI Fruit Theft Lawsuit," uttered a statement in a California courtroom that shocked everyone:
Musk once believed that the future control of OpenAI could be passed on to his children.
Wow, with one sentence, this long-running drama among OpenAI's founding team has shifted from a "corporate governance dispute" to an AI version of "Succession."
Hello everyone, welcome to Week Three of the trial: Musk vs. the OpenAI Brothers (Altman and Brockman).
Today, Altman himself testified for the first time.
In recent years, a relatively mainstream narrative has surrounded OpenAI: that OpenAI is becoming increasingly commercialized, more like a super AI company; Altman is increasingly acting like a capital manipulator; and (regardless of motive) Musk is the one who left angrily and later reported OpenAI for "betraying its original mission."
But in this trial, Altman attempted to completely reframe this story.
In his account, OpenAI is not the organization that betrayed its idealism.
From the very beginning, the person who wanted to control OpenAI and monopolize power was Musk.
Altman's First Full Account: Why OpenAI and Musk Parted Ways
The feud between Musk and OpenAI has been ongoing for quite some time, argued in the media, on social platforms, and now in court.
This trial is almost the first time Altman has stood from his perspective to give the outside world a taste of the early internal power struggles at OpenAI.
According to him, from its founding, OpenAI firmly believed and executed the principle that "AGI should not be controlled by any single individual."
To prevent super AI from being monopolized by a few in the future, OpenAI adopted a non-profit structure at its inception.
But how fickle humans are!
According to Altman's description, as time went on, Musk increasingly desired greater control, including a higher share of equity, final decision-making power over the future organization, and dominance over OpenAI's development direction.
The most explosive part was the "pass it to the children" statement.
According to Altman, there was once an internal discussion about what would happen if the person controlling OpenAI in the future passed away.
Musk's idea at the time was, "Let's just make it hereditary. If we're gone, pass the control to our kids."
Altman stated that he was very opposed to this idea at the time.
Originally, the public found it hard to grasp something like an "OpenAI organizational structure dispute," and even grew a bit tired of the drama. But "AGI control rights being hereditary" immediately lit up the eyes of gossip enthusiasts!
Especially since Musk has long cultivated a persona of upholding ideals like "open AI, humanity's future, preventing AI from being controlled by a few."
Then Altman shot a knowing smile at Musk, who was flying his plane toward China—Buddy, nobody knew, but what you envisioned back then wasn't "OpenAI for all humanity," but "OpenAI for my family."
Besides the control issue, Altman also mentioned another key event: that Musk once wanted OpenAI to merge with Tesla.
Altman strongly opposed this at the time.
In court, Altman explained that Tesla is essentially a car company with its own commercial goals, while OpenAI carries a different mission, more focused on long-term research and future infrastructure.
If merged into Tesla, OpenAI's development direction would likely be skewed by commercial objectives.
"Musk Knew All Along OpenAI Would Move Toward a For-Profit Structure"
In this trial, Altman also vehemently denied the accusation that "OpenAI betrayed its original mission."
This accusation is essentially the core narrative Musk has used to condemn and criticize OpenAI in the past.
Musk's public stance has consistently been:
OpenAI started as a non-profit with a mission to develop AI safely for humanity; but later it gradually turned into a super AI company, deeply tied to Microsoft and profit-driven.
But Altman stated in court: "Musk didn't find out later that OpenAI would move toward a for-profit structure."
According to his testimony, Musk not only knew about the relevant discussions back then, but even supported OpenAI exploring for-profit models.
During their second meeting at Tesla headquarters, he and Musk reviewed many documents outlining the creation of a for-profit company by OpenAI. Those "term sheets" detailed how much the non-profit would contribute to the new entity and what it would receive in return, including an "economic interest" in the for-profit venture.
Altman said Musk praised this move, saying the lab desperately needed massive funding.
Reuters wrote in an article about the trial that OpenAI believes Musk filed the lawsuit mainly out of jealousy over OpenAI's success after he left, and his failure to gain control of the company.
Altman also mentioned that OpenAI has now raised a cumulative $175 billion from investors for model training and computing power.
Many founders have stated that at this stage, without huge funds and massive computing power, it's impossible to continue advancing cutting-edge AI research.
OpenAI's later shift to a for-profit structure, in his view, was more a matter of practical necessity than a betrayal of idealism.
Fearing Retaliatory Action from Musk
That day, Altman also shared many details that had never been fully disclosed before.
Much of the content redefined his relationship with Musk.
For example, he mentioned that after Musk left the OpenAI board, there was internal concern that he might take some kind of retaliatory action.
Even Shivon Zilis—a member of OpenAI's founding team and the mother of four of Musk's children—advised Altman in private communications on how to consider business proposals without "upsetting" Musk.
Altman didn't elaborate with more specifics, but the statement itself is intriguing enough.
Meanwhile, during the trial, he also commented that Musk "doesn't know how to run a good research lab".
Musk's management style might work for engineering and manufacturing, but it was ineffective at OpenAI.
In his account, Musk made some key researchers feel demoralized. He asked Brockman and Ilya to list some researchers and their achievements, rank them, and then proceeded with a management style akin to a chainsaw.
"This caused enormous, long-term damage to the organizational culture," Altman said.
This is also one of the most fundamental differences Altman wanted to highlight between OpenAI and Musk.
Musk's management style has long leaned toward an "engineering iron army" model, emphasizing speed, pressure, and results; but OpenAI's core group of researchers is, by nature, closer to an academic research organization.
Conflict between the two cultures was inevitable.
Finally, it's worth noting that many attendees observed that throughout the trial, Altman talked less about "technological ideals" and increasingly used the lenses of "organizational governance" and "practical resources" to explain matters related to OpenAI.
Altman is indeed becoming more like the CEO of a large tech organization, rather than the AGI idealist entrepreneur he was in the early days.
One More Thing
Apart from OpenAI's history, part of the trial involved the famous "Altman ouster incident" of 2023.
(BTW, Ilya testified a few days ago, stating firmly that he had no regrets about participating in Altman's removal.)
Altman stated that after being removed, he seriously considered leaving OpenAI to go to Microsoft.
But he ultimately decided to return because OpenAI was too important to him.
He said, "I would run back into a burning building to save it."
References:
[1]nytimes.com/live/2026/05/12/technology/openai-trial-sam-altman-elon-musk/this-is-sam-altmans-first-time-testifying-in-court
[2]https://www.businessinsider.com/sam-altman-faces-awkward-grilling-over-toxic-culture-of-lying-2026-5
[3]https://techcrunch.com/2026/05/12/musk-mulled-handing-openai-to-his-children-altman-testifies/
[4]https://www.wired.com/story/ilya-sutskever-testifies-musk-v-altman-trial/
This article is from the WeChat public account "QbitAI," author: Heng Yu
