What Every Founder Should Learn From the Musk Altman Trial

In December 2015, Elon Musk wrote a check. Then another. Then a few dozen more. By the time he stopped, he’d sent somewhere between $38 million and $44 million to a nonprofit called OpenAI, an organization he co-founded with Sam Altman, Greg Brockman, and nine other researchers in a San Francisco office. The mission was simple on paper: build artificial general intelligence and give it away. No shareholders. No profit motive. Just open-source AI for the benefit of humanity.

Ten years later, Musk is sitting in an Oakland courtroom asking a nine-person jury to award him $134 billion and force Altman out of the company he helped create. OpenAI is now valued at $852 billion. It generates $2 billion a month in revenue. And the nonprofit charter that Musk funded? It’s been restructured, reinterpreted, and, in Musk’s words, “looted.”

The Musk v. Altman trial, which began April 28, 2026, isn’t just a legal fight between two billionaires. It’s the most expensive co-founder dispute in history, and it exposes something every early-stage founder needs to confront: what are your founding commitments actually worth if they aren’t formalized in writing?

Last updated: April 2026

Quick answers

What Musk and Altman actually agreed to in 2015

The founding story of OpenAI reads like a cautionary tale that any startup lawyer would use as a slide deck example. Musk testified that the idea started with a dinner conversation about Google. He told the jury he’d gotten into an argument with Google co-founder Larry Page, who called Musk a “speciesist for being pro-human” when Musk raised concerns about AI safety. That conversation convinced Musk that a counterweight to Google’s AI dominance was necessary.

So Musk recruited. He brought in Altman, then the president of Y Combinator. He brought in Brockman, Ilya Sutskever, and eight other researchers. Together, they launched OpenAI in December 2015 as a 501(c)(3) nonprofit. The founding group pledged $1 billion. Musk personally contributed $38-44 million in the first few years.

The founding charter was explicit: OpenAI would develop “open source technology for the public benefit” and was “not organized for the private gain of any person.” That language matters because it’s now Exhibit A in a $134 billion lawsuit.

But here’s what the charter didn’t specify: what happens if the founders disagree about how to fund the mission. What happens if one founder leaves. What happens if the nonprofit structure becomes commercially unviable. What happens if the technology they build turns out to be worth hundreds of billions of dollars.

The answer, as it turns out, is a courtroom.

How did OpenAI go from nonprofit to $852 billion company?

OpenAI’s structural pivot happened in March 2019, when the organization created OpenAI LP, a “capped-profit” subsidiary. The reasoning was straightforward: building frontier AI models required billions of dollars in compute, and no donor was going to fund that at nonprofit scale. The 2019 restructuring let outside investors put money in, with returns capped at 100x their investment. Anything beyond that cap would flow back to the nonprofit.

Musk had already left OpenAI’s board in February 2018. His departure is where the two sides’ stories diverge sharply. Musk says he left because Altman and Brockman rejected his proposal to fold OpenAI into Tesla, where Musk believed the AI research would have better access to computing resources. OpenAI’s legal team says Musk wanted total control as CEO, and when Altman and Brockman said no, he walked.

What’s undisputed is the timeline after Musk left. Microsoft invested $1 billion in OpenAI in 2019. ChatGPT launched in November 2022 and hit 100 million users in two months. By 2024, OpenAI was raising money at a $300 billion valuation. In March 2026, SoftBank co-led a round valuing the company at $852 billion. The 100x cap on investor returns effectively became irrelevant because the numbers had grown so large.

Meanwhile, OpenAI announced plans to convert fully to a for-profit corporation, though it reversed course in May 2025 after public backlash, saying the nonprofit would “remain in control” of business operations. The structural details of that control are now being litigated.

Musk filed his lawsuit in 2024, the same year he launched xAI, his own AI company that directly competes with OpenAI. OpenAI’s attorney William Savitt told the jury that Musk’s lawsuit is “competitive interference dressed as principled objection.” Musk’s legal team calls it accountability.

Why the $134 billion number matters for every startup

Musk isn’t asking for his $44 million back. He’s asking for $134 billion in “wrongful gains,” which would represent the value OpenAI created after abandoning its nonprofit mission. That legal theory, if it succeeds, would establish a precedent that founders who fund mission-driven organizations can claw back profits when the mission changes.

For founders, this creates a practical question: if you start a company with a stated mission, can a co-founder who leaves sue for the full value of the company if the mission later evolves?

The answer depends entirely on what’s in your founding documents. Musk’s problem is that OpenAI’s charter was a nonprofit mission statement, not a shareholder agreement with enforceable governance provisions. It said what the organization intended to do. It didn’t specify the consequences if it stopped doing that.

Compare that to a standard venture-backed startup. When Benchmark sued Travis Kalanick at Uber in 2017, the dispute centered on a specific shareholders’ agreement with defined board seats, voting rights, and removal provisions. The agreement was imperfect, but it was precise enough that a court could interpret it. Musk’s case requires the jury to decide what “for the public benefit” means when $852 billion is on the table.

Eduardo Saverin faced a similar problem at Facebook in 2005. Mark Zuckerberg diluted Saverin’s shares from 30% to under 10%, and internal messages later revealed Zuckerberg had written “I’m just going to cut him out and then settle with him.” Saverin eventually settled for roughly 4-5% of the company, worth billions. But the fight only happened because the original founder agreement was informal enough to allow the dilution in the first place.

The pattern repeats across the industry. Vague agreements create expensive lawsuits.

What is mission drift and how does it kill startups?

Mission drift is when a company gradually abandons its founding purpose, usually because financial incentives pull it in a different direction. OpenAI’s trajectory is the most dramatic example in tech: an organization founded to give away AI for free is now a for-profit company that charges $200/month for its premium product and is preparing for an IPO.

But mission drift doesn’t require malice. It usually happens incrementally. A nonprofit food bank starts offering career counseling to address root causes of hunger, then realizes the counseling generates more grant funding than the food distribution, then slowly reallocates staff and resources until the food bank is a career center with a food pantry in the back.

For startups, mission drift typically starts when revenue pressure forces decisions that conflict with the original vision. The company that launched to “democratize education” starts selling enterprise contracts to Fortune 500 training departments. The open-source project that promised free tools forever introduces a paid tier, then a premium tier, then deprecates the free version.

The governance mechanism that prevents mission drift is a mission-lock clause: a legally binding provision that requires a supermajority vote (typically 75-80% of the board or shareholders) to change the company’s core mission. Some organizations go further with “golden share” structures, where a mission-guardian entity holds a single share with veto power over mission changes.

OpenAI had a version of this. The nonprofit board was supposed to retain ultimate control over the for-profit subsidiary. But when the board tried to exercise that control in November 2023 by firing Altman, the for-profit side’s employees and investors revolted, and Altman was reinstated within five days. The governance structure existed on paper. It collapsed under commercial pressure.

What should be in a co-founder agreement?

The Musk-Altman trial isn’t about whether co-founder agreements are important. Every startup lawyer will tell you that. The trial is about what happens when the agreement is missing or incomplete. Here’s what the OpenAI case teaches about what founders should formalize before writing the first line of code or taking the first dollar.

Equity and vesting. Who owns what, and what happens if someone leaves? Standard four-year vesting with a one-year cliff exists for a reason: it prevents a co-founder from walking away early and keeping a large equity stake. Musk contributed $44 million but had no equity because OpenAI was a nonprofit. In a standard startup, that contribution would have been tied to ownership percentages with clear vesting terms.

Decision rights and governance. Who makes which decisions? Define board composition, voting thresholds for major decisions (raising money, changing the business model, selling the company), and what constitutes a “major decision.” OpenAI’s board structure gave the nonprofit theoretical control over the for-profit subsidiary, but the lines of authority were unclear enough that the November 2023 crisis happened.

Mission-lock provisions. If the company has a stated mission, what does it take to change it? A simple majority vote? A supermajority? Unanimous consent? OpenAI’s founding charter said “for the public benefit” but didn’t define who decides what that means or what happens if the company drifts from it.

Departure terms. What happens when a co-founder leaves, voluntarily or involuntarily? Does their equity vest? Do they retain board seats? Can they compete? Musk left OpenAI’s board in 2018, founded competitor xAI in 2023, and filed suit in 2024. A non-compete or non-solicitation clause, standard in most co-founder agreements, would have changed the entire dynamic.

Dispute resolution. Arbitration, mediation, or litigation? The fact that Musk and Altman are in front of a jury instead of an arbitrator tells you that no binding dispute resolution mechanism existed. Arbitration is private, faster, and cheaper. Jury trials produce headlines, unpredictable verdicts, and four-week courtroom dramas.

How to protect your founding vision from day one

The practical takeaway from the Musk-Altman trial isn’t “get a lawyer,” although you should. It’s that founding commitments need to be specific, written, and enforceable before the stakes get high enough to fight over. Here’s the checklist.

First, draft a co-founder agreement before incorporation. Not after. Not when you raise your seed round. Before. The six essential contracts every startup needs include a founder agreement as the first priority. Penn Carey Law School’s Entrepreneurship Clinic publishes a free founders’ agreement template that covers the basics: equity splits, vesting, IP assignment, roles, and departure terms.

Second, define your mission in legally enforceable terms. “We’re building AI for the benefit of humanity” is a press release. “The organization’s primary activity will be the development and open-source distribution of artificial intelligence research, and any change to this primary activity requires a 75% supermajority vote of the board” is a governance provision. One decorates a website. The other holds up in court.

Third, build governance structures that survive commercial pressure. OpenAI’s nonprofit board was supposed to control the for-profit subsidiary. When the board tried to exercise that control, it took five days for the commercial interests to override the governance structure. If your mission-lock clause can be unwound by employee and investor pressure, it isn’t a mission-lock clause. It’s a suggestion.

Fourth, plan for success, not just survival. Most co-founder agreements are designed for the failure case: what happens if we run out of money, if someone wants to leave, if the company folds. The OpenAI case shows you also need terms for the success case: what happens if the company becomes worth $852 billion and the original mission becomes commercially inconvenient?

Fifth, revisit your agreement at every major milestone. Raising a Series A? Review the co-founder agreement. Changing the business model? Review it again. Converting from nonprofit to for-profit? That’s when you especially need everyone at the table, with lawyers, agreeing to new terms in writing. OpenAI’s 2019 restructuring happened without Musk’s involvement because he’d already left the board. Whether that was legally sufficient is now a question for the jury.

Founders who want to survive co-founder breakups need to build the parachute before the plane takes off. The Musk-Altman trial is what happens when you don’t.

What happens next in the OpenAI trial

The trial is scheduled to run four weeks. Musk finished his initial testimony on day 3 (April 29, 2026), and cross-examination continues. Sam Altman is expected to testify, along with Greg Brockman, Microsoft CEO Satya Nadella, and several key OpenAI researchers and engineers.

The jury will decide whether OpenAI’s structural changes constituted a breach of the nonprofit’s founding charter and whether Musk is entitled to any of the $134 billion he’s seeking. Legal experts are divided on the outcome. Musk’s case depends on proving that Altman and Brockman made specific promises about the nonprofit structure and then broke them. OpenAI’s defense rests on showing that Musk himself pushed for a for-profit model and only objected after he lost control.

Regardless of the verdict, the trial has already changed the conversation. Startup lawyers across Silicon Valley report increased demand for co-founder agreements, mission-lock clauses, and governance reviews. Y Combinator, which Altman ran before joining OpenAI full-time, has long emphasized finding the right co-founder, but the emphasis has now shifted from compatibility to legal protection.

For the broader AI industry, the trial raises questions about whether nonprofit-origin AI labs can credibly commit to public-benefit missions once venture capital enters the picture. If Musk wins, it could make nonprofit-to-for-profit conversions legally radioactive. If OpenAI wins, it essentially confirms that mission statements are aspirational, not binding, unless the documents say otherwise.

Either way, the lesson for founders is the same. Put it in writing. Make it specific. Get everyone to sign it. Because by the time you’re in front of a jury, the handshake deal you remember and the handshake deal your co-founder remembers are going to be two very different stories.