The man who helped birth the AI revolution is now in a San Francisco courtroom arguing he got robbed of it.
The Summary
- Musk testified for a second day in his civil trial against OpenAI, telling the jury he texted Sam Altman "What the hell is going on?" after Microsoft invested $10 billion in early 2023
- He claims OpenAI's leaders wanted to "get rich" off a project meant to serve the public good, seeking damages for alleged "unjust enrichment to the tune of hundreds of billions of dollars"
- Altman was in the courtroom as Musk testified that Microsoft's investment signaled they expected a return, meaning effective control over AGI development
The Signal
The courtroom drama is spectacle, but the core claim cuts to the bone of Web4's foundational question: who controls the infrastructure that controls everything else? Musk's argument hinges on OpenAI's pivot from nonprofit research lab to for-profit entity backed by one of the world's largest tech companies. He cofounded OpenAI with Altman over a decade ago, and the organization's original charter emphasized building AGI for humanity's benefit, not shareholders' returns.
When Microsoft cut that $10 billion check in early 2023, Musk saw the endgame clearly. No company deploys that kind of capital without expecting control and returns. His text to Altman called it a "bait-and-switch", and his courtroom testimony spelled out why: whoever controls AGI controls the next economy. If Microsoft holds the reins, the most transformative technology since electricity gets optimized for Azure subscriptions, not human flourishing.
"Microsoft would only put $10 billion into something if they feel like they will get a return."
The trial exposes a tension that matters beyond billionaire grievances. OpenAI went from research nonprofit to $80+ billion valuation in less than a decade. Musk spent hours describing how he lost trust in OpenAI's leadership over their motives. The implication: the people building the agent layer of Web4 stopped caring about the mission and started caring about the exit.
This matters because we are watching the formation of the new stack in real time:
- Microsoft controls compute and deployment through Azure
- OpenAI controls the model layer and API access
- Developers build agents on top, but the foundation is locked to a profit-maximizing entity
Musk is not wrong that this looks different than the original pitch. OpenAI's charter talked about widely distributing benefits and avoiding concentration of power. Instead, we got the most concentrated AI capability in history, wrapped in a corporate partnership optimized for enterprise licensing.
The Implication
Watch what happens with this case. If Musk wins, it sets a precedent that founding documents and mission statements actually bind organizations, even as they scale. If he loses, it confirms what many already suspect: the AI economy will be built by the same power structures as Web2, just faster and with better PR about "responsibility."
For anyone building in the agent economy, the lesson is simple. Infrastructure matters more than applications. If you are building agents on proprietary APIs controlled by profit-maximizing entities, you are building on rented land. The ones who will own Web4 are the ones who control the compute, the models, and the economic rails underneath. Everyone else is a tenant.