It began at the Rosewood Hotel on Sand Hill Road in Silicon Valley. Late 2015. Sam Altman was thirty years old and running Y Combinator, the most powerful startup accelerator in the world. Elon Musk was forty four, already running Tesla and SpaceX, and increasingly worried about one thing: artificial intelligence.
They shared a concern. Google had just acquired DeepMind for $500 million. Demis Hassabis was becoming the most powerful person in AI. If one company controlled the most powerful technology ever created, the consequences could be catastrophic.
Their solution was simple. Build a non-profit research lab that would develop AI safely and share it with the world. No profit motive. No corporate control. Open research for the benefit of humanity.
They called it OpenAI.
Eleven years later, on Tuesday April 28, 2026, the two men sat in the same room again. This time it was a federal courtroom in Oakland, California. There was a judge, nine jurors, and lawyers on both sides. Musk was on the witness stand. Altman walked out before he testified.
They are not friends anymore.
This is the complete story. From the dinner that started it. To the courtroom that will end it. Every step in between, told as accurately and as fairly as it can be told.
No spin. No side. Just the timeline. You decide who is right.
2015: The dinner that started everything
On December 11, 2015, OpenAI was officially founded. The co-founders included Altman, Musk, Greg Brockman, Ilya Sutskever, and several top researchers recruited from Google and academia. A group of backers including Peter Thiel, Reid Hoffman, and Amazon Web Services pledged $1 billion to fund it (TIME).
Musk pushed for the billion-dollar number. In an email to the founding team, he wrote that they needed to go with a much bigger number than $100 million to avoid sounding hopeless. He said he would cover whatever anyone else did not provide.
"We need to go with a much bigger number than $100M to avoid sounding hopeless. I will cover whatever anyone else doesn't provide."
Elon Musk, founding email, 2015.
At the time, both men believed they were doing something noble. Altman expected a decades-long project. Musk framed it as protecting civilization from a technology that could also destroy it.
They were friends. They were aligned. And they had no idea what was coming.
2016 to 2017: The first cracks appear
OpenAI's early days were exciting but unfocused. The team experimented with reinforcement learning, built AI that could play video games, and published research openly. But by early 2017, a hard truth emerged.
Building AI at the frontier required an enormous amount of computing power. And computing power cost money. A lot of money. The computational resources needed for breakthroughs were doubling every three months. The non-profit model could not keep up.
Musk began to worry. According to internal emails released during the lawsuit, he told the founders that OpenAI had fallen behind Google. The gap was growing. And he had a proposal: let him take over OpenAI and run it himself. Or merge it with Tesla, where he could provide the resources it needed (TIME).
Altman, Brockman, and the other founders refused. They did not want one person controlling the organization. That was the entire point of OpenAI. No single company or individual should have that power.
Musk was not happy.
2018: The breakup
On February 21, 2018, Elon Musk officially left OpenAI's board of directors.
The public story was clean. Musk cited a potential conflict of interest with Tesla, which was developing its own AI for self-driving cars. OpenAI and Tesla were competing for the same AI talent (CNBC).
The private story was messier. Multiple reports described a power struggle. Musk wanted control. The board said no. Musk walked away.
And then something important happened. Despite promising to continue funding OpenAI, Musk slowed his donations. According to court testimony delivered yesterday, he contributed approximately $38 million in total (CNBC). That is a fraction of the $1 billion he had originally pledged to cover.
OpenAI was now a non-profit with a mission to build the most expensive technology in history and no billionaire backer to fund it.
2019: The pivot that changed everything
In March 2019, OpenAI made the decision that would define the next seven years of conflict. It created a for-profit subsidiary called OpenAI LP.
The structure was unusual. OpenAI LP would operate as a capped-profit company, meaning investors could earn returns, but those returns were capped at one hundred times their investment. The non-profit parent would still technically control the mission.
The reasoning was straightforward. Building AI required billions of dollars. No non-profit could raise that kind of money. A for-profit structure could attract venture capital and corporate partnerships.
That same year, Microsoft invested $1 billion in OpenAI. Then another $2 billion over the following years. Then in January 2023, Microsoft invested $10 billion more. By 2025, Microsoft owned roughly a quarter of OpenAI (NPR).
From the outside, this looked like a betrayal of everything OpenAI was founded to do. A non-profit created to keep AI safe and open had become a for-profit company funded by one of the largest corporations in the world.
From the inside, OpenAI argued it was the only way to survive. You cannot build frontier AI on donations.
Musk, watching from the outside, began to criticize OpenAI publicly. He called it ClosedAI. He said it had abandoned its mission. He accused Altman of prioritizing profit over safety.
The friendship was over.
2022: ChatGPT changes the world
On November 30, 2022, OpenAI launched ChatGPT. Within five days, it had one million users. Within two months, it had one hundred million. It became the fastest-growing consumer application in history.
Suddenly, the non-profit research lab that Musk helped create was the most important technology company on earth. And Musk had nothing to do with it.
In July 2023, Musk founded xAI, his own artificial intelligence company, and launched Grok, a chatbot designed to compete directly with ChatGPT. He positioned it as a free-speech alternative, uncensored and unfiltered.
Then, in November 2023, something no one expected happened. OpenAI's board fired Sam Altman.
November 2023: The five days that shook Silicon Valley
On Friday, November 17, 2023, OpenAI's board removed Sam Altman as CEO. They cited a lack of consistent candor in his communications with the board. Mira Murati, the chief technology officer, was named interim CEO.
The reaction was explosive. Within seventy two hours, more than seven hundred OpenAI employees signed an open letter threatening to quit and follow Altman to Microsoft. Microsoft, which had invested $13 billion, was blindsided. The tech industry watched in disbelief.
On Tuesday November 21, the board capitulated. Altman was reinstated as CEO. The board was restructured. Altman came back stronger than before. The people who tried to remove him were gone. Ilya Sutskever, OpenAI's co-founder and chief scientist who had played a key role in the firing, eventually left the company.
That single weekend revealed two things. First, Altman's power inside OpenAI was almost absolute. He could rebuild the company elsewhere in seventy two hours if he had to. Second, the non-profit board structure was effectively a fiction. Real control belonged to whoever had the loyalty of the engineers and the support of Microsoft.
Musk watched all of this from the outside.
"I hear you and it is certainly not my intention to be hurtful, for which I apologize, but the fate of civilization is at stake."
Elon Musk to Sam Altman, 2023 email.
2024: The lawsuit
On February 29, 2024, Elon Musk filed a lawsuit against OpenAI and Sam Altman. The accusation: they had shifted focus from public benefit to profit maximization, betraying the founding mission (TIME).
OpenAI dismissed the suit as incoherent and frivolous. In a public blog post, OpenAI revealed internal emails showing that Musk had supported the creation of a for-profit structure and had proposed merging OpenAI with Tesla under his own control.
In June 2024, Musk dropped the case. No public explanation.
Two months later, in August 2024, he refiled in federal court with stronger allegations: breach of contract, breach of fiduciary duty, fraud. The amount sought grew. By January 2026, his legal team specified $134 billion in wrongful gains to be returned to the OpenAI charity.
February 2025: The takeover bid that confused everyone
In February 2025, Musk and a consortium of investors submitted a $97.4 billion unsolicited bid to buy the non-profit that controls OpenAI. The offer was rejected within days. OpenAI said it was not for sale (BBC).
Altman's response on X became iconic. "No thank you. We will buy Twitter for $9.74 billion if you want."
That joke went around the world. It also did real damage to Musk's legal narrative.
If his lawsuit was about restoring the non-profit mission, why was he trying to buy the for-profit company? The takeover bid suggested he wanted to own OpenAI, not save it. OpenAI's lawyers would seize on this contradiction in court.
Through 2025, the public exchange escalated. Musk called Altman "Scam Altman" repeatedly on X. Altman called the lawsuits "incoherent" and "frivolous." Musk's rival xAI launched Grok, raised at a reported valuation in the hundreds of billions, and merged with X.
The rivalry was no longer personal. It was financial. Whoever won this fight would likely dominate the most important industry of the twenty first century.
April 28, 2026: The courtroom
This week, eleven years after that dinner at the Rosewood Hotel, Elon Musk and Sam Altman walked into the same room again. This time it was a federal courtroom in Oakland, California, with Judge Yvonne Gonzalez Rogers presiding.
Musk is seeking $134 billion in damages. He wants OpenAI to return to a non-profit structure. He wants Altman and Brockman removed from the board. He wants Microsoft to disgorge tens of billions in what his lawyers call ill-gotten gains.
His attorney, Steve Molo, opened with one sentence. "Ladies and gentlemen, we are here today because the defendants in this case misappropriated a charitable organization." (NPR
OpenAI's lawyer, William Savitt, responded with a different story. "We are here because Mr. Musk did not get his way with OpenAI." (ABC7) He said Musk used his promises of funding to bully founding members. He said Musk wanted to take control and merge OpenAI with Tesla. He said the other founders refused because they did not want to turn the keys of artificial intelligence over to one person. Then he said: "Musk quit, saying they would fail for sure. But my clients had the nerve to go on and succeed without him."
Then Musk took the witness stand. He was sworn in as the first witness in his own case. He told the jury he came up with the idea, the name, recruited the key people, taught them everything he knew, and provided all the initial funding. He said he specifically chose to make it something for the benefit of all humanity.
He testified about why he started OpenAI in the first place. He said the idea emerged after a disagreement about AI safety with Larry Page, Google's co-founder, who Musk claimed labeled him "a speciesist for being pro-human." He said he was worried Page was not prioritizing AI safety, which is why he wanted a non-profit, open-source alternative (CNBC). "I could have initiated it as a for-profit venture, but I opted not to," Musk said on the stand.
He warned the jury about the stakes of AI itself. He said the technology could also kill us all. Then he added: "Here we are in 2026. AI is very smart."
He told the jury that if the verdict comes out that it is okay to loot a charity, charitable giving in America will be destroyed.
OpenAI's lawyers pushed back hard. They told the jury that while other co-founders put in sweat equity, Musk showed up every few weeks to give advice and occasionally yelled at people for not moving fast enough. They pointed out that even after Musk left, Altman continued to update him on fundraising. Musk never raised concerns at the time.
Microsoft's attorney Russell Cohen argued that Musk's lawsuit had passed the statute of limitations, citing a September 2020 X post in which Musk himself wrote that "OpenAI is essentially captured by Microsoft." If Musk knew the relationship in 2020, why did he wait until 2024 to sue (CNBC)?
Judge Gonzalez Rogers scolded Musk for posting about the trial on X during proceedings. She threatened a gag order. Musk agreed to limit his posts. So did Altman and Brockman.
Altman was in the courtroom for opening statements but left before Musk took the stand. Brockman stayed and watched the entire testimony.
Toward the end of his testimony, Musk looked slightly fatigued. He was taking sips of water, rubbing his head, and running his hand through his hair. He is expected back on the stand this week.
The trial is expected to last a month. Witnesses will include Musk, Altman, Brockman, Microsoft CEO Satya Nadella, and former OpenAI chief scientist Ilya Sutskever (BBC).
"I came up with the idea, the name, recruited the key people, taught them everything I know, provided all the initial funding. I specifically chose to make it something for the benefit of all humanity."
Elon Musk, under oath, April 28, 2026.
What the legal experts are saying
Most legal observers currently view the case as harder for Musk to win on the merits (Axios). Three reasons:
- The statute of limitations is real. Musk himself acknowledged the Microsoft relationship publicly in 2020. He waited four years to sue.
- His own emails from 2017 show him proposing for-profit structures and seeking majority control of any such structure. The "betrayal" narrative does not match his historical actions.
- His $97.4 billion takeover bid undermines the claim that he is fighting to save a charity. You do not buy what you are trying to restore.
The judge has already dismissed twenty four of Musk's twenty six original claims. Only two remain in front of the jury: breach of charitable trust and unjust enrichment. The jury's verdict will be advisory. Judge Gonzalez Rogers will make the final ruling.
But trials are unpredictable. Juries are unpredictable. And the discovery process may yet reveal evidence that changes everything.
What is really at stake
This trial is not just about two men who used to be friends. It is about who gets to control the technology that will reshape the world.
If Musk wins, OpenAI could be forced to return to a non-profit structure. Its $852 billion valuation could collapse. Its planned IPO could be derailed. Altman and Brockman could be removed. And Musk's own xAI company would gain a massive advantage in the AI race.
If OpenAI wins, the for-profit model is validated. OpenAI proceeds to its IPO. Altman's position is secured. And the precedent is set: you can start as a non-profit with public promises about safety and open research, transition to a for-profit funded by the largest corporations in the world, and face no legal consequences for the change.
Both outcomes have uncomfortable implications.
The deeper question
Strip away the personalities, the dollar figures, the courtroom drama, and what is left?
A genuinely hard question. One that affects every person reading this article.
When OpenAI was founded in 2015, the founders made a promise. AI would be built openly, safely, for humanity. Not for profit. Not for shareholders. For everyone.
That promise is now in pieces. OpenAI is closed source. It is profit-driven. It serves Microsoft's interests as much as humanity's. Whether you blame Altman, Musk, the board, or just the realities of building frontier AI, the original promise is broken.
Musk's case asks: was that breach legal?
The deeper question is: was it inevitable?
Maybe AI cannot be built without billions of dollars. Maybe billions of dollars cannot be raised without for-profit structures. Maybe for-profit structures cannot be controlled by mission-driven non-profits. Maybe the entire 2015 vision was naive from the start.
Or maybe Altman and Brockman saw an opportunity to extract billions from a charity, and took it. Maybe Musk's lawsuit, however self-interested, is the only thing standing between AI and total corporate capture.
Both possibilities are real. The trial will not settle the deeper question. It will only settle the legal one.
The court will decide who is legally right. History will decide who was morally right. But the question that matters most to you is not who wins this fight. It is what happens to AI after the fight is over.
Because regardless of who wins in Oakland, the technology keeps advancing. Every day, the models get smarter. Every day, the stakes get higher. And the question that Musk and Altman asked at that dinner in 2015 is still the most important question in the world.
Who should control the most powerful technology ever created? And how do we make sure it benefits everyone?
They agreed on the answer once. They cannot agree anymore.
And neither can the rest of us.
The trial continues through May 2026. If you found this useful, share it with someone who only knows the headlines but not the full story.
