The Musk vs. OpenAI trial for an alleged breach of contract is actually an exposé of how the Silicon Valley elite is a tight network of associates, even within rival companies. The recent Pentagon agreement with seven of those companies shows how this elite is penetrating the state apparatus and transforming it.
In the case, the plaintiffs are Elon Musk and Shivon Zilis, who is the mother of four of Musk’s children whom they are co-parenting. But of course, she’s not a plaintiff because of that, but because she was a board member of OpenAI at the time. The defendants are Sam Altman and Greg Brockman, respectively CEO and President of OpenAI and cofounders with Musk. Altman is gay and married to his husband; Brockman married his wife on a workday at the OpenAI office in a ceremony officiated by Ilya Sutskever, who is also a cofounder and was the chief scientist.
You might think those personal details are irrelevant, but I would argue they are not, because at its core this trial is a conflict between two members of the same elite group. This is a group that attends (and dropout) the same educational institutions —Harvard, Yale, MIT—shares the same networks, and owns and works in similar companies, which, while some are rivals, are also deeply interconnected. And they share a similar worldview, even though they might politically take different stances.
Musk accuses Altman and Brockman of having broken the company’s initial nonprofit agreement and converting much of it to a for-profit enterprise. In March 2026, OpenAI conducted a record-breaking funding round, raising 122 billion and achieving a post-money valuation of $852 billion.
Elon has a point. He initially invested, according to him, about $100 million into a theoretical nonprofit structure with the aim to “create the first general AI and use it for individual empowerment—i.e., the distributed version of the future that seems the safest. More generally, safety should be a first-class requirement,” according to an initial email exchange between Musk and Altman. “The technology would be owned by the foundation and used ‘for the good of the world’,” wrote Altman.
The initial idea, according to the documents revealed as part of this case, was to create an AI lab that could develop technology that could eventually lead to AGI with the aim that when that happened, it would not be a proprietary technology owned by any particular individual or company—especially Google, whose DeepMind project features prominently in these initial exchanges as the competitor to beat.
That would seem almost a noble endeavor. That is, assuming that AGI was possible with current LLMs, which is highly unlikely or downright impossible, and assuming that their intentions were true, which is almost as unlikely. As the project took form, personal motivations started to show. Brockman wanted to be a billionaire. Musk wanted a majority stake that would allow him control. Altman—well, his motivations are not clear from the documents; however, from very early on, he seemed convinced that it would not be possible to achieve the company’s objective without going private.
As I understand it, Musk feels cheated by one of the greatest salesmen alive—perhaps second only to Netanyahu—and outmaneuvered. It is also important that Musk’s subsequent AI project at X was acquired by his company SpaceX and is preparing for a massive IPO. Altman is also preparing an IPO for OpenAI. He argues that Musk’s case is intended to undermine it and capture a greater share of investment. He, probably, also has a point. Nat wrote an excellent recap of the background of this trial and what is at stake.
However, what I found revealing going through the evidence presented for this case is something entirely different. Something that most of us know but to which we rarely have access: the personal connections and relationships between top CEOs, founders, and investors in the technology field. They know each other, they talk to each other, and they ask favors from each other.
For example, Musk personally asked Nvidia’s CEO, Jensen Huang, to make sure that OpenAI would get one of the first supercomputers from the firm. He also spoke to Satya Nadella, Microsoft’s CEO—who is a witness in the case because of Microsoft’s investment in OpenAI—to ensure 10,000 servers with the latest Nvidia GPUs. There are also conversations between Musk and Meta’s CEO, Mark Zuckerberg, about bidding together on the OpenAI IP.
Altman was president of the venture capital firm Y Combinator, of which Peter Thiel was an investor. The exhibits also show his contacts with Bill Gates or Pierre Omidyar, founder of eBay, as well as other tech executives.
The substance and form of those communications are both personal and professional. They are conducted through email, text message, calls, and in-person meetings which happened in offices but also in personal residences. The pattern that emerges is that of an interconnected group of individuals who control the technology that underpins much of current life in developed nations—in this case primarily Western ones—and who are moving to gain control over financial rails and defense systems.
The Pentagon has reached agreements with seven of these artificial intelligence (AI) companies: SpaceX, OpenAI, Google, Nvidia, Reflection (which doesn’t even have a public model yet), Microsoft, and Amazon Web Services. “These agreements accelerate the transformation toward establishing the United States military as an AI-first fighting force and will strengthen our warfighters’ ability to maintain decision superiority across all domains of warfare,” the Pentagon said in a statement.
The US Department of Defense—or War—has already requested $54 billion for the development of autonomous weapons which will go to some of these companies, or others, such as Palantir or Anduril, and has stated that it will request more funding for programs related to intelligence, classified and unclassified information networks (i.e., surveillance), and more.
In case you thought that Anthropic was different because they refused the Pentagon’s conditions to be able to use their technology for mass surveillance or fully autonomous weapons, Anthropic and SpaceX just signed a deal for the former to use some of the latter’s data center capacity, which begins setting the contour of alliances.
The Pentagon contracts will enmesh tech companies into the military-industrial complex. As warfare is moving away from expensive mechanical weapons systems towards smaller and cheaper AI-enabled ones, and as those same mechanical weapons introduce a ledger of AI software, technology companies will eventually co-opt the military-industrial complex.
Financial systems based on traditional banking, which introduced a ledger between issuers of money (governments) and the population, will also be eventually merged with digital AI-enabled technologies, as CBDCs or stablecoins become the standard unit of account—and control. A similar process is happening within the government systems, exemplified by DOGE, which could be understood as a pilot program. Eventually, all the important rails of the current social system will have introduced a ledger based on AI-enabled digital technologies.
This ledger will be under the control, in the West, of those very same people that are today fighting in court. This, too, might end up having a digital AI ledger. Consider their new project: objection.ai, which calls itself “the tribunal of truth” and boasts investment from Peter Thiel and other “leading VCs” on the homepage.
Supposedly, this new platform “gives everyone a fast, affordable, evidence-based way to dispute statements in the media.” The core idea is that if you feel like a publication has defamed you, you may “file an objection” with the platform, which will ask the publisher to upload evidence and will assign its own investigator to gather all evidence “regardless of whose interest it serves.” Then it will feed this information to an AI model that will “review all evidence and issue an impartial, evidence-based judgment.”
It is not difficult to see how this could become a prototype for an AI-based judicial system and how that could turn out terribly. The fact that these people are currently fighting their battle at court is a sign that they are not yet there. At the moment, they are being “invited” by the traditional power holders—the Pentagon, the financial system, the military complex, etc.—because they promise to generate more money and more power. But if how their technologies have penetrated and transformed civil life is to be an example, then we can expect something similar happening within the power grid.
However, as they become part of the power structure, their differences will have to be ironed out and resolved. We should expect more of this. If we stop seeing it publicly, then we should know why.
