A judge just greenlit testimony that Microsoft might owe Elon Musk $25 billion for allegedly breaking the rules when OpenAI went from nonprofit to profit machine.

The Summary

  • A judge ruled that Musk's damages expert can testify that Microsoft owes him up to $25 billion in his breach of charitable trust lawsuit against OpenAI and Microsoft
  • Economist C. Paul Wazzan from Berkeley Research Group calculated damages based on OpenAI's nonprofit-to-profit transformation
  • This isn't just billionaire drama, it's a legal stress test of how we handle AI infrastructure ownership when the stakes got real

The Signal

The legal machinery around OpenAI's structure just got very expensive and very real. Economist C. Paul Wazzan will testify that Microsoft's relationship with OpenAI violated charitable trust obligations, with damages potentially hitting $25 billion. That number isn't pulled from thin air. It reflects what happened when OpenAI stopped being a nonprofit AI safety project and became the computational engine behind Microsoft's entire AI strategy.

Here's what matters: Musk co-founded OpenAI as a nonprofit in 2015, poured in early capital, then watched it morph into a capped-profit entity with Microsoft as the primary beneficiary of a $13 billion investment. The legal argument is that this transformation violated the original charitable purpose. Whether you think Musk has standing or not, the testimony creates a public record of how much value transferred when OpenAI's governance changed.

This case is writing the rulebook for AI infrastructure ownership in real time. We're watching courts figure out what happens when a research nonprofit becomes the foundation of a trillion-dollar company's product strategy. The $25 billion figure puts a number on that transformation, and it's a number that makes every AI lab's board composition and investor structure suddenly very interesting.

The bigger pattern: as AI agents become economic infrastructure, the corporate structures that control them will face legal scrutiny that Web2 platforms mostly avoided. You can't just pivot from "beneficial for humanity" to "beneficial for shareholders" when you're building the operating system for autonomous work.

The Implication

Watch how other AI labs structure their governance. This case is a warning flare. If you're building agent infrastructure or have nonprofit roots anywhere in your cap table, get very clear on what obligations you actually have versus what you told early supporters. The courts are now willing to assign dollar values to mission drift, and those numbers are big enough to matter even in this industry.

Source: The Information