OpenAI’s Day of Reckoning in Oakland

In Oakland, a rivalry over artificial intelligence reaches its sharpest point yet

Sam Altman took the witness stand this week in federal court in Oakland and offered a blunt account of Elon Musk’s role in OpenAI’s early years: not as a sidelined idealist watching a nonprofit drift from its mission, but as a would-be power center who repeatedly sought control of the organization.

The testimony, delivered near the close of a closely watched trial, marked a pivotal moment in the lawsuit Musk brought against Altman and OpenAI, accusing them of betraying the company’s founding purpose by transforming a nonprofit research lab into a commercial artificial-intelligence powerhouse. By Wednesday, both sides had rested in the liability phase of the case, setting up closing arguments on Thursday and bringing the dispute to one of its most consequential junctures yet.

The case has become more than a personal feud between two of Silicon Valley’s most prominent figures. It now stands as a test of how far courts may go in policing the evolution of A.I. organizations that begin with public-interest ambitions and later pursue vast amounts of private capital.

Altman’s defense: Musk wanted the reins

On the stand, Altman rejected Musk’s claim that he had been deceived into helping build OpenAI under false pretenses. Instead, Altman described Musk as fixated on controlling the company, recounting proposals that, in OpenAI’s telling, would have handed Musk dominant ownership or influence over its future.

Altman said Musk had advanced ideas that company lawyers portrayed as alarming evidence of his approach to governance, including one suggestion that OpenAI could effectively be passed on through Musk’s family. The defense has tried to use such episodes to reframe the case: not as a story of a founder betrayed, but as one of a powerful former insider angry that he failed to gain command of a strategic asset.

Musk’s lawyers, for their part, pressed Altman on his credibility and his financial relationships, seeking to portray him as a leader who privately pursued commercial gain while publicly invoking OpenAI’s original mission. They drew on testimony from former OpenAI figures and challenged Altman’s version of how the company changed over time.

The clash laid bare the trial’s central conflict. Musk says OpenAI abandoned the nonprofit, public-benefit principles on which it was founded in 2015. OpenAI argues that Musk’s lawsuit is less about founding ideals than about competitive resentment from a rival now building his own A.I. company.

A fight over mission, money and power

At stake is not just narrative advantage. Musk is seeking sweeping remedies, including Altman’s removal, the redistribution of roughly $134 billion to OpenAI’s nonprofit side and the unwinding of the company’s for-profit conversion.

Those demands underscore how unusual the case is. OpenAI began as a nonprofit dedicated to ensuring advanced artificial intelligence would benefit humanity broadly. But the immense cost of building frontier A.I. systems pushed it toward a more commercial structure, including a deep partnership with Microsoft, which has supplied major financing and computing infrastructure.

That evolution has become a defining pattern in the industry. Labs that begin with safety-focused or public-minded missions often find themselves confronting the extraordinary capital needs of training and deploying large A.I. models. The question hanging over this case is whether a court will conclude that such a transition amounted to a legal betrayal of OpenAI’s founding commitments — or simply the practical reality of competing in one of the world’s most expensive technological races.

Why the case matters beyond the two men

The legal fight comes at a sensitive moment for the A.I. industry. OpenAI remains one of the field’s central players, and Altman one of its most influential executives. Musk, meanwhile, has emerged not only as a critic of OpenAI but as a direct competitor with his own A.I. ambitions.

That has given the proceedings an edge that extends beyond corporate governance. If Musk were to prevail in any significant way, the case could reshape how A.I. labs structure themselves, how nonprofit missions are interpreted once commercial subsidiaries enter the picture and how much authority founders retain after they leave.

A ruling favoring OpenAI, by contrast, could reinforce the broad latitude of A.I. organizations to adapt their structures as they seek funding, scale and market relevance — even when those changes draw accusations that lofty original principles have been diluted.

The court has already divided the trial into phases, with the current stage focused on liability before an advisory jury. If needed, a separate remedies phase is expected to begin around May 18 and would be handled by the judge. That structure reflects another uncertainty in the case: even if Musk can persuade jurors that OpenAI strayed from its commitments, it remains unclear whether the court will accept that he is in a position to obtain the sweeping relief he seeks on behalf of the nonprofit.

What comes next

With the evidence now closed, the case turns to closing arguments and then to the jury’s assessment of sharply conflicting accounts of OpenAI’s past.

Jurors will have to weigh whether Musk proved any actionable breach, whether Altman’s account of Musk’s desire for control is more credible than Musk’s account of institutional betrayal, and whether OpenAI’s transformation was a lawful adaptation or a violation of its original compact.

For months, the lawsuit has served as a proxy battle over the soul of artificial intelligence: whether organizations that promise to build the technology for humanity can remain faithful to that pledge once billions of dollars and strategic advantage are at stake.

This week in Oakland, that abstract debate narrowed into something more concrete — two men, two histories and one courtroom trying to decide what, exactly, OpenAI was supposed to become.

Sources

Further reading and reporting used to add context:

AI News