Musk's OpenAI trial exposes disputes

The fight here is not really about hurt feelings between Elon Musk and Sam Altman. It is about corporate control — who gets to steer a lab that started as a nonprofit safety project and became one of the most valuable companies in AI. In federal court in Oakland this week, Musk told jurors that OpenAI’s leaders effectively “stole a charity.” But OpenAI’s lawyers pushed back hard, and some of Musk’s own testimony exposed weak spots in his case. (bloomberg.com) ### What is Musk actually suing over? Musk co-founded OpenAI in 2015 and says he backed it because it was supposed to be a nonprofit research effort built for the public good, not a vehicle for private enrichment. The trial is now focused on two core claims — unjust enrichment and breach of charitable trust — aft(bloomberg.com)an and Greg Brockman from leadership, and direct any recovery back to the nonprofit side. (thenextweb.com) ### Why is the nonprofit part so important? Because Musk’s whole theory depends on mission, not contract. He says OpenAI’s leaders promised one thing in public and built something else in private. One piece of evidence getting a lot of attention is a 2017 Brockman diary entry saying that if OpenAI committed to nonprofit status and then quickly pursu(thenextweb.com)ample evidence for the claims to be heard at trial, which is a big reason this case got this far. (thenextweb.com) ### Where did Musk’s case get shaky? The simplest problem is that Musk acknowledged there was no written contract governing the terms of his early donation. Under cross-examination, he said he contributed based on his understanding of OpenAI’s mission and charter. That does not kill the case — charitable-trust arguments can reach beyond a normal con(thenextweb.com)bout intent, governance, and whether nonprofit commitments were real or strategic. (bloomberg.com) ### Why did xAI come up? Because OpenAI wants to show Musk is not just a wounded founder — he is now a direct competitor with his own AI company. Trial coverage also surfaced an awkward point for Musk: he confirmed that xAI had at least partly used OpenAI models to help train Grok through model distillation. Th(bloomberg.com)xAI. It gives OpenAI a way to frame him as conflicted, not purely principled. (theverge.com) ### What can the court actually do? A lot, at least in theory. Musk is seeking remedies that could include unwinding the restructuring and forcing leadership changes. One report pegged the potential damages claim at as much as $150 billion, though even before trial the judge had questioned parts of Musk’s valuation logic. Also, the jury is advisory here on key issues — the judge will ultimately deci(theverge.com)t the real power still sits with the bench. (thenextweb.com) ### Why does this matter beyond OpenAI? Because every frontier AI lab now has the same tension baked in. Safety missions sound nonprofit. Training giant models costs for-profit money. OpenAI is just the clearest test case of what happens when those two logics collide. If Musk wins meaningful relief, founders, donors, and state regulators may get a s(thenextweb.com)t lofty founding language does not do much once capital, compute, and commercial scale take over. (thenextweb.com) ### So what is the bottom line? This trial is exposing something bigger than one Silicon Valley feud. It is showing how flimsy “for humanity” promises can look once an AI lab becomes strategically important and wildly valuable. Musk may or may not win. But the testimony already made one thing clear — the argument over who AI should serve was never settled inside OpenAI, and now a federal court is being asked to sort it out. (thenextweb.com)