Before the bell rings

Part 4

ON APRIL 7, 20 DAYS BEFORE TRIAL, Elon Musk dropped a bombshell. He said that he wasn’t looking for any money for himself.

It was a surprise because in January he had filed a “Notice of Remedies” that said, “the primary monetary remedy that Plaintiff seeks is disgorgement of wrongful gains.” In other words, he wanted the defendants to cough up all the value (potentially $134 billion) that they had gotten unfairly.

The January notice did not say to whom the wrongful gains would be paid, but it was easy to assume that it would be Musk.

The new filing was called an “Amended Notice of Remedies,” and it proclaimed that Musk’s central purpose in suing had always been “to prevent the subordination of a public charity … to private, for-profit interests.”

Musk charged that despite “solemn promises that OpenAI would operate as a nonprofit for the benefit of humanity,” and having “accumulated billions of dollars in assets under that charitable shelter, [defendants] converted those assets into a wealth machine for themselves, Microsoft, and Silicon Valley insiders.”

The goal of the lawsuit, Musk said, was to put a stop to that.

To that end, therefore he wasn’t asking that the $134 billion in wrongful gains be paid to him. The money would be directed to “the OpenAI charity,” though how that would work wasn’t explained. He wasn’t even asking to get back the money that he actually donated.

He went on to say that if he proves that Sam Altman and the other defendants are liable for breaching the charitable trust, he will also ask the court to impose a suite of “equitable remedies.”

This was also a surprise. It was late in the case to be bringing in new issues and would be sure to generate a fierce response from OpenAI.

But the surprise went well beyond the timing of the original notice of remedies, Musk said — in what seemed an afterthought — that in addition to the money, he would also ask for “equitable relief, including an injunction,” but that would be something for the court after the trial, and no details were described.

But now his newly proposed remedies were direct and aggressive. He said the court should remove Altman as CEO and director of OpenAI, as well as require the enterprise to honor its original charitable undertaking to operate for the benefit of humanity and not for the financial benefit of any individual.

The judge should require Altman and co-founder Gregory Brockman to give up their financial benefits from OpenAI and compel Microsoft — accused of aiding and abetting the breach of OpenAI’s charitable mission — to give up its allegedly ill-gotten gains.

Finally, in the most far reaching of the proposed remedies, Musk called for the restoration of the company’s original nonprofit structure and the complete unwinding of the for-profit enterprise that currently operates OpenAI’s commercial development of artificial intelligence.

Such a structural change would likely undo the company’s October 2025 restructuring previously greenlighted by the attorneys general of California and Delaware, the states with regulatory authority over OpenAI’s charitable mission.

In sum, he was saying he did not want anything for himself, and if the defendants were liable, the judge should do what was needed to restore the original concept that this enterprise needed to function, in reality — not just on paper — for the benefit of humanity.

The filing said that if what Musk wanted was not clear before, “it should be clear now.”

Framed this way, Musk was asking U.S. District Court Judge Yvonne Gonzalez Rodgers to construct equitable relief to accomplish the goal he said he had always sought.

But the strategy was not without risk.

Equitable remedies are decided by the judge, not the jury, and a federal judge has wide discretion in deciding what things are appropriate. What she would determine to do might not be anything at all like what Musk wanted. This was the judge, after all, who had denied his preliminary injunction in part because she did not want to hamstring competition in the fluid and fast-moving world of AI development.

With his filing he was putting the centerpiece of his claim in her hands, without knowing what would happen.

In poker terms, Musk was going all-in.

OpenAI responds

The OpenAI team was not very happy with Musk’s last-minute pivot, and they didn’t delay in letting Gonzalez Rogers know.

On April 10, they filed a response that accused Musk of a “legal ambush.” They said that while Musk purported to “clarify” the remedies he was seeking, he was actually trying to change his entire position. OpenAI accused Musk of directly contradicting statements he had been making to the judge “for months,” and was now seeking relief that he had previously rejected. They alleged he was “sandbagging” (the poker term for a player who raises after passing) and intentionally injecting chaos into the litigation, all to the end of recasting the public narrative.

The filing boiled it down this way: “Musk has abandoned the only remedy he has pursued. He cannot now replace it on the eve of trial with extravagant new remedies that are legally improper and factually unsupported. Without a remedy, there remains no case left to try.”

Rogers Gonzalez called a hearing for April 17 and directed the parties to meet immediately and discuss if the case should be “bifurcated” — that is, heard in two parts. The first would focus on whether Musk had shown that OpenAI and Microsoft were liable. If Musk was successful, then there would be a second stage that would consider damages, both the disgorgement of unjust gains and the newly surfaced equitable claims.

When the parties arrived at the April 17 hearing, they said they were agreeable to bifurcation, though how it would work was subject to extended discussion with the judge.

OpenAI’s lawyers wanted an opportunity to challenge Musk’s right to raise equitable remedies. They said that the relief he was seeking would require evidence that had not been part of discovery. For example, in order for a court to issue an injunction, the judge is required to balance the harm to the defendant against the harm to plaintiff and to the public interest. That would presumably involve testimony about how OpenAI as a company, its employees, customers and the public would be affected. According to OpenAI that complex topic had not been prepared for trial because it hadn’t been part of Musk’s case.

Gonzalez Rogers did not make a ruling on the timeliness of the remedies but ordered briefing and said she would “address the availability of Musk’s proposed remedies at a later date.”

Of course, you can’t get any remedy unless the defendants are found to be liable. So If the first phase of the trial did not result in a finding of liability, there would be no remedies at all.

And if the jury found that Musk filed the suit too late, the judge would likely agree and end the case without further action.

As to what the jury would know about the remedies Musk was requesting, the judge was emphatic. She said the lawyers would not be allowed to say anything to the jury about remedies during the damage phase. She repeated herself several times to add to the gravity of the point. And after issuing her written order, she doubled down, “During this phase, the parties, their counsel and witnesses shall not present, elicit, argue, suggest or otherwise refer to any specific remedy or form of relief that Musk requests.”

For Musk, that meant the jury would decide liability without knowing that his proposed remedies would include directing the $134 billion to the “OpenAI charity” and unwinding OpenAI’s for-profit structure. To the extent that his goal was to show the jury his “selflessness” in pursuing the lawsuit, the judge’s ruling seemed to have thwarted it.

Who will win?

And where does this all leave us on April 28 when the jury is empaneled and the shooting starts? Will it be Musk, the world’s least sympathetic plaintiff, or Altman, the world’s most conniving defendant, who comes out on top?

Not surprisingly, public interest in the trial has generated plenty of speculation, and in our age of crowd-sourced wisdom, the prediction markets are there to gauge the sentiment of the wagering class.

The Polymarket prediction market is taking bets on the question “Will Elon Musk win his case against Sam Altman?”

As of Thursday, April 23, 2026, at 5:00 p.m., the betting is 35% in favor of Musk on a volume of $108,333.

In other words, Altman is a 2-1 favorite.

Musk’s chance of winning has ranged from a high of 61% on Jan. 14, 2026, to 22% on Feb. 27, 2026, according to Polymarket, so while not good, his odds have improved in the last two months.

Musk is doing better over on the Kalshi prediction market. On April 23, Musk was given a 40.1% chance to prevail on the question “Will Elon win his case against OpenAI?”

A total of $$374,968 has been wagered. Musk’s odds have ranged from 68.2% on Jan. 16, 2026, to a low of 31.8% on March 10, 2026.

(The rules of decision and the precise questions posed are similar but not the same on the two sites.)

The prediction markets largely focus on who will be the winner after a trial, but if one were to go by the odds in normal federal litigation, the proper bet would be that Musk and Altman would settle at the last minute.

Most federal court cases settle, and though there are ongoing academic debates about how to determine the exact percentage, whatever the methodology, more than 50% of federal civil cases settle.

And in this situation, there are many good reasons for each party to settle. OpenAI has a lot to work with at trial, but if its trial case goes south, it faces a worst-case outcome. The 2025 restructuring would be unwound, its desired initial public stock offering impossible, and there could be a new judge-created corporate structure that would be unlikely to include Sam Altman. Is it likely? Probably not, but a big risk nonetheless.

And Musk? The statute of limitations is a big issue for his claims. There were many events that might have tipped him off to the idea that his original vision of a humanity-facing nonprofit was being unraveled. Did he know before Aug. 5, 2020, that there was a fox in the chicken coop? And if he didn’t, should he have known? This was the world’s richest man — a man who not only had unlimited resources but was perceived to be one of the smartest people on the planet. Is it possible he was just a mushroom kept in the dark?

A veteran court watcher would argue this case needed to settle. Even the cockiest bantam roosters would have to realize that in this game of chicken, they could end up fricassee.

And yet, predicting a settlement would be a chancy bet. It would take Musk’s agreement to make a settlement. And what did he have to lose? He wasn’t trying to recover any money. If he loses, he will remain the richest man on the planet. Moreover, his x.AI is a competitor to OpenAI. If he loses, he can still appeal; if he wins, make it harder for OpenAI to raise capital before the appellate court weighs in?

Moreover, there is a favorable fact that might help him on the key statute of limitations question.

Musk began his funding of OpenAI in May 2016. Many things happened after that date that might have signaled that there had been a breach of charitable trust, but Musk continued to make donations to OpenAI. In fact, His donations continued until Sept. 14, 2020, five weeks after the Aug. 5, 2020 limitations date.

At trial Musk can argue that if he really knew the nonprofit dream was being snuffed out before Aug. 5, 2020, he would have stopped giving OpenAI money. But he didn’t.

Based on all that, there is a non-trivial chance Musk will push his case all the way to the very end.

We shall see.