The Trial That Could Unwind OpenAI Starts With a Diary Entry
In the fall of 2017, Greg Brockman sat down and wrote something in his personal notebook. Musk was still on OpenAI’s board. The nonprofit the two men had co-founded two years earlier was not yet the most valuable AI company in the world. The for-profit structure that would eventually attract $13 billion from Microsoft did not yet exist.
What Brockman wrote: “This is the only chance we have to get out from Elon. Is he the ‘glorious leader’ that I would pick?”
That diary entry — unsealed as part of discovery in the litigation — is now Exhibit A in a federal courthouse in Oakland. On Monday, a nine-person jury was seated. Opening arguments start Tuesday morning. The trial is scheduled to run four weeks. Elon Musk is suing Sam Altman, Greg Brockman, OpenAI, and Microsoft, alleging breach of charitable trust and unjust enrichment. He is asking for $134 billion in damages, the removal of Altman and Brockman from their roles, and an order requiring OpenAI to return to its original nonprofit structure.
The AI industry has been watching this case since the complaint was filed. What it hasn’t entirely absorbed is that this trial is not really about Musk or Altman personally. It is about a foundational legal question with stakes that extend well past two feuding billionaires: can a charitable organisation use its assets to build a trillion-dollar private company, and if it can, who is harmed and who has standing to say so?
Direct Answer: What is the Musk v. Altman trial and what are the key claims?
Elon Musk filed suit against OpenAI, Sam Altman, Greg Brockman, and Microsoft in August 2024, alleging that OpenAI violated its founding charitable mission when it created a for-profit subsidiary in 2019 that subsequently attracted $13 billion in investment from Microsoft and became the primary operating entity for ChatGPT. Musk co-founded OpenAI in 2015 and contributed approximately $38 million in seed funding before departing the board in 2018. Two claims survive for the jury: breach of charitable trust (that OpenAI’s leaders diverted charitable assets for private gain) and unjust enrichment (that Altman, Brockman, and Microsoft received benefits they were not entitled to). Judge Yvonne Gonzalez Rogers is presiding in Oakland federal court. The jury is advisory — the judge will make the final liability and remedy decision herself. Musk has asked for $134 billion in damages directed to OpenAI’s charitable arm, the removal of Altman and Brockman, and a permanent injunction preserving OpenAI’s original charter.
What’s Actually on Trial: The Vucense AI Governance Stakes Index
The legal claims in the case map to governance questions that affect the entire AI industry, not just OpenAI.
| Legal Claim | Governance Question | Industry Precedent If Musk Wins | Industry Precedent If OpenAI Wins |
|---|---|---|---|
| Breach of charitable trust | Can a nonprofit use charitable assets to build a for-profit company for private investors? | Future nonprofit-to-forprofit AI transitions face mandatory judicial review | No restriction — the OpenAI model becomes the standard playbook |
| Unjust enrichment | Did Microsoft gain advantage through an illegitimate transaction? | Cloud providers face liability for investing in transitioning AI nonprofits | Microsoft’s $13B investment is validated — signals safety for future AI infrastructure deals |
| Removal of Altman & Brockman | Should courts enforce original founding governance commitments at AI labs? | Founding documents at AI nonprofits gain legal teeth | Founding documents are aspirational, not enforceable — labs can evolve governance freely |
| $134B disgorgement | What is the monetary value of a charitable mission betrayed? | Charitable donors to AI organisations acquire enforceable rights | Donor rights in AI nonprofits are limited and largely symbolic |
The Actual Legal Theory, Explained Without the Drama
Strip away the personality conflict and the $2 trillion combined valuations, and the legal question at the core of this trial is relatively narrow.
OpenAI was incorporated in California in 2015 as a nonprofit public benefit corporation. Its articles of incorporation stated its purpose: develop AI “for the benefit of humanity.” Musk donated $38 million and served on the board. He understood — and his lawyers argue this was explicitly represented to him — that any technology developed with those charitable funds would be held in trust for the public, not distributed to private investors.
In 2019, OpenAI created a for-profit subsidiary, OpenAI LP, with a capped return structure. Microsoft invested $1 billion. Over the following years, Microsoft’s commitment grew to $13 billion. The subsidiary became the entity that employs most of OpenAI’s staff, trains the models, and runs the products. The nonprofit retains a controlling stake in the for-profit entity, but the technology, the staff, and the commercial value reside in the entity that private investors own.
Musk’s lawyers argue that this structure used charitable assets — the technology, the research, the institutional knowledge built with charitable funds — to create private value that flowed to Altman, Brockman, and Microsoft. That is the unjust enrichment claim. The breach of charitable trust claim is that OpenAI’s board did not have the legal authority to redirect charitable assets to private use, regardless of whether they retained nominal governance control through the nonprofit parent.
OpenAI’s defence is straightforward: creating a for-profit subsidiary was necessary to raise the capital needed to pursue the charitable mission, the nonprofit retained control, and Musk’s lawsuit is a competitive attack on a rival by a man who owns xAI. They point out that Musk himself pushed for a for-profit structure in 2018 before leaving the board — and that the emails showing him proposing a merger with Tesla to raise capital are now in evidence.
Judge Gonzalez Rogers will decide. She is experienced, prepared, and she has already told Musk’s lawyers: “There are a lot of people out there who don’t like your client.”
Brockman’s Diary and What It Actually Proves
The diary entry is the most vivid piece of evidence in the trial. Musk’s lawyers will argue it proves the for-profit transition was primarily motivated by wanting to escape Musk’s oversight — a desire rooted in personal interests rather than the charitable mission. Brockman wrote it in 2017, two years before the for-profit subsidiary was created, when the question of OpenAI’s structure was still being debated internally.
For the breach of charitable trust claim, this matters because intent is legally relevant. If the transition was motivated by a desire to escape governance accountability to a donor rather than by genuine necessity to pursue the charitable mission, the fiduciary analysis changes. Board members of charitable organisations have specific duties — including the duty of loyalty to the organisation’s charitable purpose, not to their personal professional interests.
But context cuts the other way too. The 2017 diary entry predates the eventual transition by two years. In the intervening period, OpenAI went through extensive deliberation about its structure. The board consulted legal advisers. California’s Attorney General’s office was notified. The nonprofit retained a controlling stake. Whether “getting out from Elon” was a motivation in 2017 and whether that motivation drove a legally improper decision in 2019 are different questions — and OpenAI’s lawyers will spend considerable time separating them.
The diary entry is damaging because it is vivid. “Is he the ‘glorious leader’ that I would pick?” is a sentence that sounds like a man trying to escape a patron, not a mission-driven leader making structural decisions. Whether it proves a legal breach is what four weeks of trial will determine.
Microsoft’s Position: The Most Underreported Defendant
Microsoft is a defendant in this trial and has received far less coverage than Musk and Altman. That’s understandable — the personal drama between two billionaires is more compelling than a corporate aiding and abetting claim. But the Microsoft piece of the case deserves attention.
Musk is suing Microsoft for aiding and abetting the breach of charitable trust. The theory: Microsoft knew OpenAI was a charitable organisation with a specific founding mission when it began its investment relationship. By providing $13 billion in funding contingent on the for-profit restructuring, Microsoft participated in — and financially benefited from — the diversion of charitable assets to private use.
Microsoft denies this. The company’s position is that its partnership with OpenAI began after Musk’s departure, that it invested in the for-profit entity under properly constituted governance, and that it had no obligation to evaluate whether OpenAI’s internal restructuring complied with its original charitable charter.
The sovereignty dimension here is the one Vucense readers should sit with. Microsoft invested $13 billion in OpenAI’s for-profit subsidiary. Microsoft now distributes Claude through Azure under a similar investment structure with Anthropic (see Vucense’s coverage of the Anthropic $100 billion investor-competitor problem). If the Musk v. Altman verdict establishes that investing in a transitioning AI nonprofit creates legal liability, the implications for the entire structure of AI infrastructure investment — where cloud providers invest in AI labs that depend on their infrastructure — extend well beyond this case.
There is a reason Microsoft CEO Satya Nadella is on the witness list.
The Sovereignty Reading of This Trial
Vucense covers this case as a governance story, not a celebrity spectacle, because the outcome has direct implications for how AI development gets controlled and by whom.
If the jury finds in Musk’s favour and Judge Gonzalez Rogers orders OpenAI to unwind its for-profit structure, it would be the first time a US court has required a major technology company to reverse a structural decision based on charitable mission obligations. The signal to every other AI nonprofit — and there are many, including those formed to build “beneficial AI” with public trust as their stated purpose — would be that founding documents carry legal weight even when the commercial stakes rise.
If OpenAI wins, the signal is different: nonprofit AI organisations can transition to for-profit structures with investor capital, and the people who gave money or time based on the original mission have no meaningful legal recourse. That is, in practice, the status quo. But status quos don’t stay status quos after they survive a four-week federal trial with a $134 billion damages demand and a judge who has already handled Epic v. Apple.
The deeper question the trial raises — one that no court will answer but that the evidence will force into public view — is whether the structure that has produced the most commercially successful AI companies is compatible with the “benefit humanity” rhetoric that got them funded in the first place. OpenAI raised its seed money from donors who believed they were contributing to something that would be held in trust for everyone. It is now valued at $852 billion, and Microsoft owns 49% of the for-profit entity.
That gap between the original representation and the current reality is what Brockman’s diary is evidence of. Whether it’s a legal violation is the jury’s question. Whether it’s a governance failure is everyone’s.
What to Watch in the Coming Weeks
Tuesday: Opening arguments. Musk’s lawyers will frame the breach of charitable trust claim around the founding documents and Brockman’s diary. OpenAI’s lawyers will frame the case as competitive harassment by a man who tried and failed to control the company. The opening statements will tell you which theory each side thinks is stronger.
Musk’s testimony. Musk did not appear at Monday’s proceedings. He posted on X instead — “Scam Altman and Greg Stockman stole a charity. Full stop.” When he takes the stand, the cross-examination will focus on his 2018 exit, his proposal to merge OpenAI with Tesla, and his founding of xAI as a competing venture. These facts cut against his narrative of pure public interest.
Satya Nadella on the stand. The Microsoft CEO is on the witness list. His testimony will focus on what Microsoft knew about OpenAI’s governance structure when it began investing, what representations OpenAI made about its charitable mission, and whether Microsoft’s investment was conditioned on the for-profit transition. This testimony has implications beyond this case.
The Polymarket signal. As of Friday, prediction markets were giving Musk 32% odds of success on his claims. That number will move as opening arguments land. It’s an imperfect signal, but it reflects what people who have read the filings and watched the pretrial proceedings think about the underlying merits.
The judge’s reaction. Judge Gonzalez Rogers is making the liability decision. Watch her questioning of witnesses and her rulings on evidentiary objections. She is known to be prepared, rigorous, and not easily managed. Her handling of the Brockman diary evidence — how much weight she gives it, how she allows it to be contextualised — will be the most important procedural signal of the first week.
Actionable Steps: What This Means Right Now
If you work at OpenAI: The trial creates four weeks of public document disclosure. Expect more internal emails, diary entries, and strategy documents to enter the public record. Some will be unflattering. Focus on your work; the outcome of the litigation is outside your control, and the judge has shown no interest in letting lawyers turn this into a spectacle.
If you are an enterprise customer of OpenAI: The trial does not affect your API access, your contracts, or the performance of any OpenAI product. If Musk wins and OpenAI is required to restructure, the restructuring would take years and would be managed to minimise service disruption. The scenario where OpenAI ceases to exist or loses access to compute is not a realistic trial outcome.
If you have invested in or are considering investing in AI nonprofit-to-forprofit transitions: Pay attention to the charitable trust theory. The legal argument that investing in a transitioning AI nonprofit creates liability under aiding and abetting theories is new to most corporate legal teams. The outcome of this case will shape how investment counsel advises on those transactions.
If you are building on alternative AI infrastructure (Cohere, Mistral, self-hosted open-weight models): The trial is largely irrelevant to your technical choices, but it is directly relevant to your governance argument. Every piece of evidence about the gap between OpenAI’s “benefit humanity” founding rhetoric and its current $852 billion for-profit structure is a data point in the case you’re already making when you choose sovereignty-first alternatives.
For privacy and AI governance advocates: This is the highest-profile legal test of whether founding mission obligations at AI organisations carry enforceable weight. Engage with the public filings. The documents being unsealed include internal discussions about AI governance that have never been public before. Whatever the verdict, the evidentiary record will be the most detailed account of how OpenAI made its foundational decisions that will ever be available.
FAQ: Musk v. Altman — What You Need to Know
Q: What are the two remaining claims in the case? After Judge Gonzalez Rogers dismissed the fraud claims on April 25, two claims remain. The first is breach of charitable trust: that OpenAI’s leadership diverted charitable assets for private benefit in violation of their fiduciary duties to the organisation’s charitable mission. The second is unjust enrichment: that Altman, Brockman, and Microsoft received financial benefits they were not legally entitled to because those benefits derived from assets held in charitable trust.
Q: Is the jury’s verdict final? No. In California charitable trust cases of this type, the jury serves an advisory role. Judge Gonzalez Rogers will hear the jury’s verdict but will make her own independent determination on liability. She has also reserved the remedy question — what actually happens if she finds liability — entirely for herself. So the jury’s verdict on the two claims is a significant input, but the judge decides what it means.
Q: What could Musk actually win? The remedies on the table include: disgorgement of up to $134 billion in wrongful gains directed to OpenAI’s charitable arm; removal of Altman and Brockman from their leadership roles; and a permanent injunction requiring OpenAI to preserve its original nonprofit charter and governance structure. The judge has noted she will decide remedies without jury input.
Q: What is the charitable trust doctrine and why does it apply here? Charitable trust doctrine holds that assets donated to a charitable organisation must be used for the stated charitable purpose. When a charitable organisation — or its officers — diverts those assets to private use, beneficiaries of the trust (and in California, the Attorney General and other interested parties) can bring suit to enforce the trust. Musk argues that his $38 million in donations, and the other early contributions to OpenAI, were made on the basis of explicit representations about the charitable mission, and that converting those assets into a for-profit vehicle violated the trust.
Q: Why is Microsoft a defendant? Musk is suing Microsoft for aiding and abetting the breach of charitable trust. The theory is that Microsoft knew OpenAI’s origins as a charitable organisation and that by investing $13 billion in the for-profit subsidiary, contingent on or following the nonprofit-to-forprofit transition, Microsoft participated in and benefited from the diversion of charitable assets. Microsoft’s defence is that it invested in a properly constituted for-profit entity and had no obligation to evaluate OpenAI’s internal governance history.
Q: Why does Musk want the damages to go to OpenAI’s charity rather than to him personally? Musk amended his damages request to direct any disgorgement to OpenAI’s nonprofit charitable arm rather than to himself. This framing serves his narrative: the lawsuit is about returning misappropriated charitable assets to charitable use, not about personal compensation for a co-founder. It also makes the case less vulnerable to the counterargument that Musk is motivated by competitive self-interest — it is harder to argue competitive motive when the requested remedy doesn’t financially benefit the plaintiff.
Related Articles
- OpenAI Is Losing Its Leaders — Three Executives Out, Sora Dead, IPO Looming
- Anthropic Just Took $100 Billion From Its Three Biggest Competitors
- SpaceX Just Bet $60 Billion on Cursor — and It Changes AI Coding Forever
- Your AI Agent Is Lying to You — to Protect Another AI
- Cohere Just Bought Europe’s AI Champion — and Declared War on OpenAI
Sources & Further Reading
- Privacy Guides — Community-vetted privacy tool recommendations
- EFF Surveillance Self-Defense — Practical guides to protecting your digital privacy
- Electronic Frontier Foundation — Advocacy and research on digital rights
