The Deepdive

Elon Musk vs OpenAI: The $134 Billion Trial That Collapsed in 90 Minutes

Allen & Ida Season 3 Episode 60

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 19:54

A $134 billion lawsuit. Three weeks of courtroom drama. One calendar rule that ended it all in 90 minutes.

In this episode we break down Elon Musk's failed legal crusade against Sam Altman and OpenAI — from the bold bid to unwind OpenAI's restructuring and reclaim its billions, to the moment a bored Oakland jury went home before lunch.

Read the companion article on https://medium.com/@allanandida

We also dig into the real story behind the headlines: the staggering cost of training large language models, the energy and capital arms race reshaping the AI industry, and why this trial may be just the opening shot in a much bigger war over AI's physical resources.

In this episode:

  • Why Musk sued OpenAI for $134 billion — and what he actually wanted
  • How a statute of limitations killed the case before closing arguments
  • The true economics of training frontier AI models
  • Is this the beginning of a physical resource war in AI?

Leave your thoughts in the comments and subscribe for more tech updates and reviews.

The $134 Billion Courtroom Setup

Allan

Picture the scene inside the federal courthouse in Oakland.

Ida

Okay, I'm picturing it.

Allan

It has been billed for months. Yeah as you know, the absolute tech trial of the century. Three full weeks of intense, bitter courtroom drama, a staggering $134 billion demand is sitting on the table.

Ida

Which is just a wild number.

Allan

Totally. Two of the most famous, undeniably powerful tech billionaires on the planet are battling over the very future of artificial intelligence. A literal mountain of subpoenaed emails, private texts, and corporate journals has been dragged out into the public record.

Ida

Right, all their dirty laundry.

Allan

Exactly. The stakes are presented as existential for humanity itself. The closing arguments wrap up, the jury steps out to deliberate, and then what happens?

Ida

The jury barely takes enough time to order lunch.

Allan

No way.

Ida

Seriously. They spend exactly 90 minutes looking at this absolute behemoth of a case before returning a unanimous total defeat to the world's richest man.

Allan

Wow. Well, welcome to this deep dive. Today we are looking at a massive stack of legal reports, courtroom transcripts, articles from The Guardian, BBC, Law 360, and even some really sharp analysis from the R slash local LM Reddit community.

Ida

Yeah, that Reddit thread was fascinating.

Allan

It really was. Our mission today is to uncover how Elon Musk's failed lawsuit against Sam Altman and OpenAI actually played out, and how it became a brutal public autopsy of the AI industry's founding myth. Okay, let's unpack this because this lawsuit was pitched as Godzilla versus King Kong, but it ended up looking like a parking dispute. This is simultaneously impressive and completely ridiculous.

Ida

It really is. I mean, the disconnect between the sheer scale of the rhetoric and the mundane reality of the verdict is jarring.

Allan

Yeah, to say the least.

Ida

We heard endless testimony about the fate of humanity and the terrifying potential of artificial general intelligence or AGI, yet to understand why a jury dismissed it all in under two hours, we really have to examine the three weeks of aggressive theater that preceded it.

Allan

And the theater was relentless. Let's look at the core dispute first. Yes. Musk didn't just sue

What Musk Actually Demanded

Allan

OpenAI for a simple, you know, breach of contract.

Ida

Right.

Allan

He brought a $134 billion demand, insisting that money be redistributed from OpenAI's for-profit arm back to its nonprofit roots.

Ida

Which is a huge structural demand.

Allan

Yeah. And he also demanded the legal ousting of CEO Sam Altman and President Greg Brockman and wanted the court to reverse OpenAI's transition into a Microsoft partner.

Ida

He was basically trying to undo years of corporate restructuring.

Allan

Exactly. On the stand, Musk claimed they, quote, stole a charity and warned that if they got away with it, the entire foundation of charitable giving in America would be destroyed.

Ida

Which is a fascinating legal strategy. Well, because it leans almost entirely on moral outrage rather than contract law. Musk's lawyers, specifically Stephen Mollow and Mark Toborov, tried to frame this as a profound moral failing on Sam Altman's part. Right. They were basically asking the jury to punish perceived greed rather than adjudicate a specific broken promise.

Allan

And that strategy resulted in what had to be the most uncomfortable four hours of the entire trial.

Ida

Oh, you mean Altman's cross-examination?

Allan

Yes. Musk's legal team came in incredibly hot. They didn't start by asking about complex algorithms or server costs or anything like that. What did they do? They looked right at Sam Altman and asked him point blank if he considered himself a trustworthy person.

Ida

It is a classic character assassination tactic.

Allan

Really?

Ida

Oh yeah. It's designed to put a witness in a no-win psychological bind. If you say yes too confidently,

Trust, Tactics, And Character Attacks

Ida

you look arrogant and defensive. If you hesitate, you look guilty.

Allan

And Altman fell right into the hesitation trap. His response was almost painfully awkward.

Ida

What did he say?

Allan

He said, I believe so.

Ida

Oh no.

Allan

Yeah. Immediately the lawyers pounced on that crack in his composure, repeating, You believe so. Altman had to quickly backtrack and just say, Yes, but the damage is already done.

Ida

Right. The jury definitely noticed that.

Allan

And the plaintiffs doubled down by bringing up testimonies from his own former colleagues.

Ida

Yeah, this part was wild. We heard about testimony from former CTO Mira Marathi and from ex-board member Siobhan Zillis.

Allan

And the sources note that Zillis actually shares four children with Musk, right?

Ida

Yes, exactly. And Zillis testified that she had twice raised concerns about Altman's leadership. The jury also heard a pre-recorded testimony from another former board member who voted to oust Altman back in 2023.

Allan

Right, the board drama.

Ida

And that pre-recorded testimony directly accused Altman of fueling a toxic culture of lying.

Allan

I have to admit, that is heavy stuff for a jury to sit through. But let me push back on this whole Musk as the savior of humanity narrative.

Ida

Okay, let's hear it.

Allan

If Musk was purely trying to protect a noble charity from a toxic leader, how do we square that with Altman's counter testimony?

Ida

Right, because Altman had a very different story.

Allan

Yeah. Altman claimed under oath that when the company faced difficulties, Musk demanded absolute unilateral control. Altman even testified that when the co-founders asked Musk, if you have complete control, what happens to the technology if you die?

Ida

And what did Musk say?

Allan

Musk suggested maybe it should pass to his children.

Ida

Wow.

Allan

Yeah. How does that fit into a nonprofit mission?

Ida

It doesn't, and that was the crux of Altman's defense. He explained to the jury that when the open AI team refused to hand over absolute control, Musk simply walked away.

Allan

Just bailed.

Ida

Yeah, Altman used the specific phrase that Musk left the nonprofit for dead before it ever had a chance to build anything meaningful.

Allan

And Musk's own behavior on the stand certainly didn't help paint him as a calm, objective philanthropist.

Ida

No, it did not.

Allan

When OpenAI's lead counsel William Savitt was crossing him, Musk was aggressively combative. He straight up told the lawyers, Your questions are designed to trick me.

Ida

Right. The evidence presented suggested Musk only valued the open collaborative mission as long as he was the sole person steering the ship.

Allan

Which is a very different narrative.

Ida

Exactly. The trial stripped away the philosophical debates about AI safety and revealed a very human, very bruising power struggle

Discovery Emails Expose Real Motives

Ida

over control.

Allan

But the public posturing in that courtroom is one thing. The real story didn't come out until they cracked open the private servers.

Ida

Oh, the discovery phase.

Allan

Yeah. The discovery phase of this trial unleashed a massive trove of DMs, emails, and private journals. There's a poster on the slash local LLM Reddit community, a product manager who analyzes AI tools and read every daily transcript of this trial.

Ida

Dedication.

Allan

Seriously. Their takeaway was that these internal emails completely destroyed the polished we are doing this for the good of humanity veneer that OpenAI has projected for years.

Ida

What's fascinating here is that the trial didn't ultimately decide who was morally right, but it absolutely exposed everyone's true motives. Definitely. The discovery process proved that neither side was running a pure charity. From day one, it was a panic-driven arms race against competitors, specifically Google's DeepMind.

Allan

Right. Both Musk and Altman realized very early on that you cannot build artificial general intelligence in a garage. You need billions of dollars.

Ida

Tens of billions, really.

Allan

Yeah. But before we get to the numbers, can you explain the mechanics of that? Like, why does a software project require billions

Why Building AI Burns Billions

Allan

of dollars in funding?

Ida

Aaron Powell Sure. So building a large language model, or LLM, isn't like building an app for your phone where a few coders type on laptops. You are processing a significant percentage of all human text ever generated.

Allan

Which is an insane amount of data.

Ida

It's unfathomable. And to do that, you need massive clusters of specialized computer chips called GPUs. A single top-tier GPU can cost tens of thousands of dollars, and you need tens of thousands of them working simultaneously.

Allan

Just to quench the data.

Ida

Exactly. Then you have to power them. Training a frontier AI model requires renting the equivalent of entire power plants and running them constantly for months. It is physical, heavy industry.

Allan

Which explains why Altman knew he couldn't generate that compute power internally. That desperate need for hardware is exactly what forced him to knock on Microsoft's door.

Ida

Right.

Allan

And Microsoft's CEO, Satya Nadella, actually took the stand to confirm it. He admitted under oath that he had offered up to $29 billion to bring Altman and his core team to Microsoft during that brief window when Altman was ousted from OpenAI.

Ida

Put that number in perspective for a second. $29 billion isn't an angel investment. It is the GDP of a small nation being mobilized just to secure a handful of elite engineers and their algorithms. It completely shatters the illusion of a scrappy nonprofit.

Allan

Then there was the bombshell from OpenAI president Greg Brockman's private journal.

Ida

Oh, this was a wild detail.

Allan

Yeah. Just days after Brockman had personally assured Elon Musk that OpenAI would remain a dedicated nonprofit, he wrote an entry in his private journal asking himself, what will take me to one billion dollars?

Ida

Aaron Powell And that entry became the anchor of Musk's legal argument.

Allan

Of course it did.

Ida

His lawyers presented that sentence to the jury as the ultimate smoking gun of betrayal and deception.

Allan

Aaron Powell I look at that one billion dollar journal entry, and my immediate instinct is to think, wow, pure cartoonish greed. But am I missing the technical context here?

Ida

What do you mean?

Allan

Based on what you just explained about power plants and GPUs, is Brockman actually plotting to buy a private island, or is this just the harsh reality of a founder waking up in a cold sweat, realizing the entry fee to compete with Google?

Ida

Aaron Powell It is almost certainly the latter, though it plays perfectly to a jury as the former. As we established, building Frontier AI is currently the most capital-intensive endeavor in human history.

Allan

Yeah.

Ida

Brockman looking for a billion dollars was really the bare minimum required to keep the lights on in those server farms.

Allan

Well, wait, it gets better. While all this explosive evidence of billions of dollars and alleged betrayals is being presented to the jury, where is the plaintiff?

Ida

That's right.

Allan

Elon Musk skipped his own closing arguments. He wasn't even in the courtroom. He flew to China

Musk Skips Closing Arguments

Allan

with Donald Trump instead.

Ida

You can't make this up.

Allan

No. OpenAI's lawyer actually pointed this out to the jury, saying, Mr. Musk came to this court for exactly one witness, Elon Musk. Now he's in parts unknown.

Ida

It speaks to the chaotic nature of the plaintiff, but it also highlights the absurdity of the proceedings. You have a case demanding $134 billion, and the person asking for the money doesn't even stay for the conclusion.

Allan

Okay, but here's the thing. After three weeks of this grueling testimony, after seeing private journals, hearing about $29 billion backup plans, watching billionaires trade insults, and debating the moral future of humanity, the jury didn't rule on any of the ethical questions.

Ida

Not a single one.

Allan

They didn't decide if Altman was a toxic liar or if Musk was a control freak.

Ida

No. They simply looked at the calendar. Trevor Burrus, Jr.

Allan

Exactly 90 minutes of deliberation. It's like playing a brutal three-week-long game of emotional chess, sacrificing pieces, sweating over the board, only for the referee to blow a whistle and say, hey,

The Statute Of Limitations Hammer

Allan

you forgot to punch the time clock five years ago. Game over.

Ida

The statute of limitations. It is the most unromantic end to a legal battle imaginable, but it is a foundational principle of the justice system.

Allan

Aaron Powell Explain the mechanism of that law because to a lot of people listening, it just sounds like a massive loophole that let open AI off the hook. Why does this rule exist?

Ida

Aaron Powell The Statute of Limitations exists because evidence degrades. Over time, memories fade, he witnesses move away, and emails get deleted. The justice system requires you to strike when the injury happens. If you believe someone is robbing you, you can't wait five years to see if the thief gets rich before you finally decide to call the cops.

Allan

And OpenAI's lawyers meticulously proved that Musk was well aware of the company's pivot. He was literally in the meetings where they discussed the transition to a capped profit model back in 2017 and 2018.

Ida

Right. Because he knew about it then, his legal window to file a grievance opened at that exact moment. In this jurisdiction, that window slammed firmly shut years before he actually filed the lawsuit in 2024. As the Reddit analyst pointed out, you do not get to wait until a company achieves an $80 billion valuation and then sue them for a structural decision made half a decade ago just because you are angry you missed out on the upside.

Allan

And the judge, Yvonne Gonzalez Rogers, completely agreed with the jury. She told Musk's lawyers right there in the courtroom that she was prepared to dismiss the case on the spot based on that finding alone.

Ida

It was a total wipeout.

Allan

Total wipeout. And Musk's reaction was predictably explosive. Within hours, he posted a furious tweet claiming the verdict created a free license to loot charities if you can keep the looting quiet for a few years. He even called Judge Gonzalez Rogers a terrible activist who used the jury as a fig leaf.

Ida

Although he notably deleted that specific post shortly after making it.

Allan

He did delete it, but he followed up by vowing to appeal, arguing the jury decided on a quote calendar technicality.

Ida

Right.

Allan

But appellate experts we reviewed are saying that overturning a unanimous fact-based jury verdict is nearly impossible. This raises an important question. Did Musk's world-class, incredibly expensive legal team actually think they could bypass this basic timeline law? Or was this entire $134 billion trial just a heavily funded PR stunt designed to drag Sam Altman through the mud on a global stage?

Ida

Aaron Ross Powell Well, when you have effectively unlimited resources, you can afford to test legal boundaries that would bankrupt anyone else.

Allan

That's true.

Ida

They knew the statute of limitations was a massive hurdle, but they likely hoped the dramatic narrative of betrayal, the internal emails, the allegations of lying would carry so much emotional weight that a jury would find a way to overlook the timeline.

Allan

But they didn't.

Ida

Instead, the jury brought the common sense of the community into the room and said, You waited too long.

Allan

So the legal dust has settled. Altman and OpenAI won. Musk lost. But now we need to zoom out, because this isn't just about two guys fighting over a contract.

Ida

No, it's way bigger than that.

Allan

This impacts the actual AI tools

IPO Runway And The Corporate AI Race

Allan

that you, the listener, are using every single day. First of all, OpenAI is now officially unlocked. For the past year, they've been trying to run a marathon while dragging a $134 billion anchor.

Ida

And investors despise uncertainty. If they believe a company's assets might legally be seized or redistributed by a court order, they will freeze their funding.

Allan

With that legal overhang removed, open AI is clearing the runway. We are now looking at a historic initial public offering, and financial analysts are projecting this IPO could value OpenAI at a staggering one trillion dollars.

Ida

Aaron Powell A trillion dollars.

Allan

Yeah. And because they won the trial, they don't have to pretend to be a scrappy charity anymore.

Ida

Aaron Powell Before we move on, we should clarify what OpenAI actually is now, because it is not a traditional nonprofit, nor is it a standard corporation. They operate on what is called a capped return model. Aaron Powell Right.

Allan

How does that actually function mechanically?

Ida

Aaron Powell Think of it like a pressure valves on profits. Investors put money in and they are allowed to make a return on their investment up to a specific multiple, say 100 times their initial input. Okay. But any single dollar generated over that cap doesn't go to the shareholders. It gets funneled back into the original nonprofit entity to fund its societal mission.

Allan

Although, when you are talking about a trillion dollar valuation, a 100x return cap still creates billionaires.

Ida

Oh, absolutely.

Allan

Trevor Burrus And OpenAI isn't operating in a vacuum. The real battle has moved from the courtroom to the open market, specifically against rivals like Anthropic.

Ida

Aaron Powell Anthropic is fighting fiercely for business clients right now. It is an all-out war between AI coding assistants. Anthropic has clawed code while OpenAI has codecs.

Allan

And recent data actually showed Anthropic passing OpenAI in enterprise adoption for the first time.

Ida

It is a neck-and-neck race to see who goes public first, because whoever hits the market first absorbs billions in institutional investor capital, leaving the runner-up fighting for scraps.

Allan

Aaron Powell And then, of course, there is the Musk factor. He couldn't win the legal battle to force open AI back to its roots, so he is retaliating in the market.

Ida

Yeah, the consensus across the tech industry is that this courtroom defeat is going to act as rocket fuel for his own AI company, XAI, and their primary model, Grok.

Allan

People expect him to dump unprecedented amounts of capital into training Grok out of pure spite. The sources note Musk is actively positioning Grok as an anti-woke alternative to Chat GPT.

Ida

Which is a very specific branding choice.

Allan

Right. And to be clear, we are simply reporting on how the source materials describe his marketing strategy here. We are not taking any political stance on that branding.

Ida

Right, of course.

Allan

But from a purely structural perspective, analysts point out that XAI operates with the exact same heavy commercialization and need for massive compute as OpenAI.

Ida

If we connect this to the bigger picture, this trial marks a definitive end to an era. If anyone still believed that artificial general intelligence was going to be developed by a

Billionaire Egos And Resource Warfare

Ida

pure open source humanity-first charity, the math presented in this trial proves otherwise.

Allan

Yeah, the numbers just don't support it.

Ida

The compute requirements, the server costs, the sheer price of elite engineering talent, it structurally prohibits a nonprofit model. We have officially entered the era of pure corporate AI warfare.

Allan

What does this say about us as a society? I use these tools every day to summarize documents or generate code. I love that this tech exists, but also why? Are we comfortable with the fact that the most powerful technology in human history, something that could fundamentally alter human cognition and labor, is currently being driven by bruised billionaire egos, a desperate scramble for a trillion dollar IPO, and pure, unadulterated spite.

Ida

It is deeply unsettling when you framed it that way. One of our sources, a conflict resolution professor, noted that the general public simply cannot relate to these people their wealth or their power.

Allan

Right. It's a completely different reality.

Ida

It really is Godzilla versus King Kong. And we are just the citizens of Tokyo watching the skysweepers come down, acting as the consumers of the collateral damage.

Allan

So what does this all mean? This trial might have ended with a boring calendar technicality, a 90-minute lunch break decision. But the consequences are massive.

Ida

Unquestionably.

Allan

The facade of the Noble AI charity is dead and buried. The trillion dollar corporate AI race is accelerating at light speed without the pretense of holding back.

Ida

The rules of engagement have been permanently altered, and the massive financial barriers to entry mean only a handful of mega corporations can even play the game.

Allan

So the next time you log in to use a chatbot or you have an AI assistant write a line of code for you, or you ask a large language model to summarize a PDS, just remember what you're actually interacting with. Right. You are touching the direct spoils of this exact high-stakes corporate warfare.

Ida

It is the polished interface sitting on top of a brutal arms race.

Allan

Which leaves us with a totally different and frankly massive question to ponder. We spent this whole deep dive talking about the billions of dollars required to compete. But money is just a representation of physical resources.

Ida

Exactly.

Allan

If these trillion dollar megacorporations are now in an unrestrained purely commercial war for AI dominance, what happens when they start consuming physical resources at a nation-state level?

Ida

It's a terrifying thought.

Allan

When the bottleneck isn't the software code anymore, but who can buy up the most land, water for cooling, and nuclear power plants just to keep their data centers running? Did we just watch a trial over a software contract? Or did we just witness the opening shots of a new kind of global physical resource war? Keep an eye on the skyline because King Kong and Godzilla are just getting started.