Breaking apart friendships isn’t simple, however few are as messy and costly because the disintegration of Elon Musk and Sam Altman’s once-thriving tech bromance, which has now come to a authorized finish.
On Monday, a jury dominated in opposition to Mr. Musk in a lawsuit in opposition to OpenAI, alleging that Mr. Altman and different executives “stole charity” (as one among Mr. Musk’s legal professionals put it) by turning most of the as soon as nonprofit analysis amenities into big firms. (Disclosure: Vox Media is one among a number of publishers that has signed a partnership settlement with OpenAI. Our reporting stays editorially unbiased.) For 3 weeks, legal professionals on either side have laid out more and more shaky proof in an try and discredit each males and show that they’re untrustworthy and power-hungry.
Musk claimed he was tricked into donating roughly $38 million to OpenAI underneath false pretenses and is suing for $150 billion in monetary damages, together with vital adjustments to OpenAI’s management and governance construction. On condition that OpenAI first added a business sector in 2018, Decide Yvonne Gonzalez Rogers accepted the jury’s resolution that Musk couldn’t sue inside the three-year statute of limitations. However the proof introduced in court docket may very well be sufficient to influence state regulators to rethink the settlement that allowed OpenAI to reorganize right into a for-profit firm within the first place.
Attorneys say Musk is more likely to attraction the decision and the catfight is probably not over but. However even past the result, the trial solid an usually uncomfortable highlight on the inside workings of Silicon Valley and the AI business. Listed below are 5 key info revealed at trial.
OpenAI board members questioned Sam Altman’s integrity
Mr. Musk’s authorized staff sought to color Mr. Altman as deeply untrustworthy, somebody vulnerable to mendacity to his co-founders, staff and board members to additional his personal pursuits.
A number of former OpenAI staff and administrators gave related testimony in court docket. Former board member Helen Toner mentioned in a video affidavit that Altman’s “sample of habits related to honesty and candor” immediately led to his short-term elimination as CEO in 2023. Mira Murati, former chief expertise officer at OpenAI, testified that he had a bent to “say the identical factor to some folks and the exact opposite to others.” In a single occasion, she mentioned Altman blatantly lied in regards to the security evaluations required to vet new AI fashions.
Greg Brockman saved a diary – and he in all probability needs he did not.
A few of the extra sordid proof introduced at trial got here from a private diary saved by OpenAI President Greg Brockman, by which he recorded his “stream of consciousness” as he thought of whether or not turning OpenAI right into a for-profit firm could be “morally bankrupt.”
“I can not think about turning this right into a business enterprise and not using a very nasty dispute,” he wrote in a 2017 entry. “It is mistaken to steal a nonprofit from him,” she mentioned, referring to Musk, who co-founded OpenAI and supplied most of its startup capital. “He is actually not silly,” Brockman later wrote. “His story will finally be that we weren’t trustworthy with him.”
Brockman was additionally candid about his private ambitions. “It will be nice to make billions,” he wrote. He then acquired a stake in OpenAI, which is at present estimated to be price round $30 billion.
Shocked, shocked: Elon Musk is tough to cooperate with
In 2017, OpenAI constructed a extremely superior bot that was in a position to beat high skilled gamers within the strategic multiplayer battle recreation Dota 2. This was a serious milestone for the up-and-coming lab. “It is time for OpenAI to take the following step, and that is the catalyst,” Musk wrote in an electronic mail to Brockman.
Mr. Musk gave Mr. Brockman and co-founder Ilya Satskeva a brand new Tesla Mannequin 3, presumably to “cheer us up,” Mr. Brockman testified. Tesla’s CEO then summoned them to a self-described “haunted home” to debate the potential for a business arm of OpenAI, the place Musk’s then-girlfriend Amber Heard provided them whiskey.
At one level, Mr. Musk turned so enraged by the visitors’ insistence that they share management of OpenAI fairly than cede absolute management to Mr. Musk, “I really thought he was going to hit me or bodily assault me,” Brockman testified. Within the months that adopted, Mr. Musk repeatedly steered that Tesla take up OpenAI, Mr. Altman testified. And in a single “notably cringeworthy second,” he determined he ought to move on OpenAI to his youngsters.
Musk ultimately left OpenAI in 2018 and commenced constructing his personal competitor. Throughout the all-hands assembly, Musk received into one other tense argument with Josh Achiam, now chief futurist at OpenAI, over the race to develop synthetic basic intelligence. “He received offended and referred to as me an fool,” Achiam testified. In recognition of Atiam’s bravery, two OpenAI staff (together with Dario Amodei, who later left to discovered Anthropic) introduced him with a small golden statuette formed like a donkey’s butt and engraved with the message, “By no means cease being a idiot for security.”
Microsoft groups up with OpenAI to keep away from being left behind within the AI race
Musk’s preliminary funding for OpenAI stemmed from the breakdown of one other friendship with Google co-founder Larry Web page. Musk mentioned that at his birthday celebration, Musk mocked him for preferring people to computer systems. Microsoft, which was named in Musk’s lawsuit for aiding and abetting OpenAI’s abandonment of its nonprofit mission, later turned OpenAI’s first main company investor in 2019, as Microsoft needed to compete with Google as AI competitors intensified.
“I do not wish to be IBM,” Microsoft CEO Satya Nadella wrote to executives, referring to the corporate’s decline within the private computing race, in line with emails revealed at trial. “It was changing into extra core and vital to have actual possession at each layer of the stack,” Nadella testified.
That meant embracing each nook of the OpenAI world. Microsoft performed a key function in bringing Mr. Altman again to energy after a failed board coup in 2023, however Mr. Nadella referred to as Mr. Altman “beginner metropolis, so far as I am involved.” In a textual content thread revealed in court docket, Altman referred to as on Microsoft executives to scrutinize numerous members of OpenAI’s reconstituted board of administrators. They now management each the for-profit firm and the previous nonprofit group.
One firm government testified that Microsoft plans to take a position greater than $100 billion in OpenAI by this summer season. Final fall, the corporate acquired a 27% stake in OpenAI.
Everybody needs to rule the world (synthetic basic intelligence)
Microsoft. Musk. Altman. Blockman. Virtually everybody who testified on the trial accused one other boogeyman of too impure motives and too corrupt a persona, and identified that trusting folks to manage what everybody agrees to is a essential expertise. In distinction, their very own self-reflection took a backseat to their ambitions.
“We do not desire a Terminator-like ending,” Musk testified, drawing a transparent eye roll from Decide Gonzalez Rogers, who tried and didn’t steer the case away from a dialogue of the existential dangers of AI. “If we’ve got somebody answerable for AI that we do not belief, I believe that is an enormous hazard to the entire world,” Musk mentioned.
Greater than a decade in the past, Musk teamed up with OpenAI’s co-founders to create a charity that was poised to tackle one other menace that was then making an attempt to steer the AI race: Google, which had not too long ago acquired Demis Hassabis’ DeepMind. Now, like Altman and Brockman, who testified that they resisted Musk’s authoritarian makes an attempt to safe absolute management over synthetic basic intelligence, Musk portrayed himself as somebody selfless and clear sufficient to take duty.
“It’s ironic that regardless of these dangers, your consumer is constructing an organization that’s on this very area,” Gonzalez-Rogers as soon as advised Musk’s legal professionals, referring to xAI, which has come underneath hearth this 12 months for facilitating the mass creation of non-consensual deepfakes. “I believe there are numerous individuals who do not wish to put the way forward for humanity in Mr. Musk’s fingers.”
Replace, Might 18, 2026, 2:00 PM ET: This text has been up to date to replicate the conclusion of the trial.
I learn 1 article final month
Right here at Vox, we’re unwavering in our dedication to protecting the problems that matter most to you: democracy, immigration, reproductive rights, the setting, and the rising menace of polarization throughout our nation.
Our mission is to offer clear, accessible journalism that permits you to keep knowledgeable and assist form your world. Changing into a Vox member immediately strengthens our means to offer in-depth, unbiased reporting that drives significant change.
We depend on readers such as you. Please be a part of us.
Swati Sharma
vox editor in chief
