• The founders walk in, and the past walks with them
• What Brockman says he does now
• Control: the original fault line
• Musk’s Tesla play—and Zilis’ emails
• The Cerebras question: disclosure or dodge?
• Journals, billion‑dollar dreams, and the for‑profit pivot
• Cracks in the philanthropic story
• Performance under fire
• OpenAI’s turn to clean up
• The tension that won’t go away

Greg Brockman Testifies in Musk v. Altman Trial Human Human coverage portrays Brockman’s testimony as undermined by his own emails and journals, highlighting inconsistencies between his professed loyalty to a philanthropic, Altman-led OpenAI and his apparent openness to lucrative for‑profit structures without Musk. It stresses questions about incomplete disclosure to Musk, the optics of his personal financial ambitions, and a courtroom demeanor that suggests defensiveness more than clarity. @Verge OpenAI’s courtroom drama over its soul took a sharp turn when its president, Greg Brockman, was forced to narrate—under oath—how a gleaming non‑profit ideal morphed into a high‑stakes fight over control, money, and Elon Musk’s place in the story.

§The founders walk in, and the past walks with them

On May 4, Greg Brockman and Sam Altman entered the courtroom together, the co‑architects of OpenAI now sitting on opposite sides of Elon Musk’s narrative about what the company was supposed to be.¹ Brockman was slated to testify immediately, positioning him as the trial’s next crucial narrator.

He began by rewinding to the beginning: the earliest days when OpenAI was “initially supposed to be part of Y Combinator, as a research arm.”² In a solicitation email to then‑Yahoo CEO Marissa Mayer, Brockman boasted that donors included Elon Musk, Jessica Livingston, Reid Hoffman, and Peter Thiel—and added, “I’m personally donating $100,000,” a pledge he never actually fulfilled.²

Outside the courtroom, the importance of Brockman to the project has never really been in doubt. Years earlier, Sam Altman had publicly gushed it was “impossible to imagine openai succeeding without greg!”³ That praise now hangs awkwardly over a trial where Musk claims OpenAI betrayed its founding promises.

§What Brockman says he does now

On direct examination, Brockman painted himself as OpenAI’s operational backbone. Asked what he does as president, he answered with a line that landed somewhere between start‑up flex and courtroom cringe: “I do all the things.”⁴

He also sketched his view of where the technology stands. Describing the current generation of AI models, Brockman said, “We very much have these AI models that are smart and capable but they’re not fully connected to the world,” adding that “we as society are still figuring out how do we integrate these.”⁵ It was a sweeping, almost visionary framing—one that contrasts sharply with the grubby paper trail the lawyers then dragged into the light.

§Control: the original fault line

To understand why Musk is suing, you have to go back to 2017, when the core dispute wasn’t yet about revenue, but about power.

Emails from that period show that “the big sticking point for Brockman and Sutskever was control.”⁶ They “didn’t want Musk — or anyone — to have control over OpenAI.”⁶ In a message to Musk lieutenant Jared Birchall, Shivon Zilis summarized their position bluntly: “You and I can argue that’s stupid all we want but they are holding firm on it.”⁶

From Musk’s camp, this looks like the hinge moment: the founders pulling away from the billionaire backer who believed he’d be steering the ship. From Brockman’s side, it looks like an early declaration that no single person—Musk included—should own a project whose stated mission was to benefit humanity, not a cap table.

§Musk’s Tesla play—and Zilis’ emails

Musk, for his part, was not sitting still. Evidence at trial shows he explored folding OpenAI into Tesla, effectively turning the non‑profit into an in‑house lab for his car company. Zilis, then an influential adviser in Musk’s orbit, pushed the idea hard.

“We are seeing more details about Zilis advocating for Musk’s plan to wrap OpenAI into Tesla,” one account of the evidence notes. “Tesla solves the funding issue immediately… Tesla at least has option to bury,” reads one email from Zilis. “They haven’t internalized the advantages to burying this in Tesla for stealth advantage,” reads another.⁷

At the same time, Musk dangled prestige in front of Altman. As part of “his push to increase Tesla’s AI presence,” Musk “offered Sam Altman a board seat at Tesla,” according to a 2017 email from Zilis noting that “those who want to work on large scale AI research don’t currently think of Tesla, and Elon wants to change that by announcing his intention to create a world-class AI lab.”⁸

Musk’s perspective is straightforward: if he was funding and championing OpenAI, pulling it into Tesla and building a top‑tier AI operation there was a natural evolution. To critics, the emails read like a blueprint to quietly absorb a public‑spirited lab into a secretive, for‑profit automaker.

§The Cerebras question: disclosure or dodge?

The trial also surfaced smaller but telling skirmishes over transparency. One key line of attack: whether Brockman properly disclosed that he owned a stake in Cerebras when OpenAI considered a merger with the chip company.

Brockman says he told Sam Teller and Shivon Zilis, “who were in theory Musk’s chiefs of staff,” but not Musk personally.⁹ Musk’s lawyer, Molo, hammered the point, suggesting that looping in staff was not the same as notifying Musk himself.⁹ The live commentary on the proceeding was unimpressed: “I am less convinced by. Lots of executives delegate.”⁹

From Musk’s camp, this detail feeds a broader story that Brockman played cute with conflicts of interest. From Brockman’s side, it’s the reality of how large organizations function: you tell the lieutenants, not always the general.

§Journals, billion‑dollar dreams, and the for‑profit pivot

The most damaging material for Brockman didn’t come from Musk’s emails but from Brockman’s own pen.

In one much‑discussed entry, Brockman wrote a line that Musk’s lawyers have turned into a symbol of corrupted ideals: “Financially what will take me to $1B?”¹⁰ Molo argued this showed a man more interested in plotting personal wealth than “figuring out funding for the non-profit.”¹⁰ Brockman insisted “there’s more context,” but the phrase is sticky, and Molo leaned on it hard.¹⁰

The context that did emerge is hardly comforting. Brockman recalled debating whether to accept “Elon’s terms,” or “reject the terms, he quits to create his own [AI company], and then we create our own [AI company].”¹⁰ That hypothetical has, in effect, come true: Musk walked away and built xAI.

It got worse. In another entry, Brockman foresaw exactly the narrative Musk is now selling to the jury: “His story will correctly be that we weren’t honest with him in the end about still wanting to do the for profit just without him.”¹¹ Six days after assuring Musk that he and others wanted “more results in the non-profit and to fundraise there,” Brockman wrote, “We’ve been thinking about that maybe we should just flip to a for-profit. making money for us sounds great and all.”¹¹

In court, Brockman shrugged this off as “an expression of a frustration and not a plan.”¹¹ Molo dryly asked if he’d rehearsed that answer. Brockman said “no.”¹¹ Either way, the entries make him look—at best—like someone who understood exactly how betrayal would look and moved forward anyway.

Another passage from the same journal cuts the other way, and Musk’s team seized on it. “Greg Brockman’s journal: ‘it’d be wrong to steal the non-profit from him,’” one summary notes, referring to Sam Altman.¹² Brockman also wrote, “To convert to a b-corp without him. That’d be pretty morally bankrupt.”¹² The commentator called it “probably the most solid thing the Musk team has gotten out during the trial so far.”¹²

If you’re Musk, this is gold: the president of OpenAI acknowledging—in his own handwriting—that the exact maneuver Musk now alleges would be “morally bankrupt.” If you’re Brockman, it’s evidence that you once wrestled seriously with the ethics and, in your telling, tried not to cross certain lines.

§Cracks in the philanthropic story

Molo also pressed Brockman on whether the non‑profit entity should still be “a philanthropic endeavor.” Brockman answered, “forcefully, ‘no.’”¹³ That was followed by diary notes quoting Musk himself as saying they had to “figure out how do we transition from non-profit or something which is essentially philanthropic endeavor and is b-corp or c-corp.”¹³

The effect, as one observer put it, “does make Brockman look pretty shifty.”¹³ Musk’s side points to that shift as proof that OpenAI abandoned its mission. OpenAI’s defenders counter that everyone involved, Musk included, saw early on that achieving artificial general intelligence would require capital and corporate structures that a pure non‑profit couldn’t sustain.

§Performance under fire

Through it all, Brockman’s demeanor became its own subplot. One reporter watching the cross‑examination noted that “Brockman is not doing himself any favors.”¹³ Another observed that while his exchanges weren’t as volcanic as Musk’s, “he’s also pushing back on a lot of questions,” with frequent lines like “I wouldn’t characterize it that way,” “I’m not sure I’d say it that way,” and requests to see statements “in context.”¹⁴ “It’s not as contentious or tense as Musk’s exchanges with Savitt, but it’s definitely notable.”¹⁴

Earlier in the day, he’d been described as speaking “very quickly and very softly,” enough that the judge had “just scolded him for it.”² In other words: not the smooth, controlled technocrat OpenAI probably wanted as its frontman, and not the explosive showman Musk turned himself into on the stand—something in between, and under visible strain.

§OpenAI’s turn to clean up

After Musk’s team finished, “OpenAI’s lawyers are now getting their shot at Brockman.”¹⁵ Observers were “curious to see what they can recover from this testimony.”¹⁵

Their task is unenviable. On one side, they have to show that a shift toward a capped‑profit structure was a rational response to technical and financial reality—one even Musk had mused about, according to Brockman’s notes.¹³ On the other, they must neutralize Brockman’s own journals, which read like a director’s commentary for Musk’s legal complaint.

§The tension that won’t go away

Strip away the courtroom theatrics and a clear chronology emerges: a star‑studded, non‑profit AI lab is founded inside the Silicon Valley trust circle; its leaders quickly realize AGI will demand far more money and control than they first imagined; Musk pushes to pull the project closer to Tesla and his own orbit; Brockman and Sutskever push back, determined that “Musk — or anyone” shouldn’t control it.⁶

From there, journal entries, side emails, and shifting corporate forms fill in the gaps. Each side now insists the other betrayed the original deal. Brockman swears he’s trying to navigate a fast‑changing reality; Musk claims the navigation route was drawn behind his back.

The trial won’t settle the question of who “should” own AGI. But Brockman’s testimony has already settled something else: whatever OpenAI began as, it didn’t stay that way for long—and everyone involved saw the turn coming.

Story coverage