Before the bell rings

Part 1

IN A JURY TRIAL that might decide the fate of all humanity, should you root for the world’s least sympathetic plaintiff or most devious defendant?

That’s a hyperbolic but directionally reasonable way to frame the question to be resolved in the litigation extravaganza set to convene in federal court in Oakland this April, the case in which Elon Musk seeks to bring down Sam Altman’s AI empire, OpenAI (the company behind ChatGPT).

On Jan. 15 and March 25, U.S. District Court Judge Yvonne Gonzalez Rodgers entered orders that appeared to clear the way for Musk to tell his story to a jury of his peers. The judge set aside four weeks for the trial, and barring last minute maneuvering by the lawyers or an eleventh-hour settlement, the slam fest is set to begin April 27.

The case has been widely dismissed as vanity litigation — another Elon Musk sideshow — but underneath the shimmering frou-frou, something serious is at stake.

As it is framed, the lawsuit is about whether the awesome power of machine intelligence will be controlled by the same tech platforms and the same billionaire bros who have corrupted the Internet, or will there be at least one frontier enterprise that is dedicated to directing that awesome power to something bigger than money, something that benefits all of us.

The stakes are high. Massive amounts of money and power and reputation are in play. There will be winners and losers. But unlike most court cases, we all have a stake in this one.

Whether it is the cost of — or climate impact caused by — massive data centers, the loss of jobs, the evisceration of privacy, the new tools for authoritarians, the new frontiers in cyberterrorism, the potential for rogue bioweaponry or the non-trivial risk of an extinction event, there are no sidelines to stand on as the development of AI accelerates. Every person on the planet will be affected. And it is happening right now, all around us.

Musk v. Altman is one of those generational law cases that straddle the fault lines of our society, and how it is decided — rightly or wrongly — will matter for years and years to come.

Before it all gets started, let’s take a look at how we got to this point.

A shared mission

Musk’s litigation journey — begun with maximum bravado in state court in San Francisco on Feb. 29, 2024 — has been littered with one setback after another.

He started out with a filing that was unusually hyperbolic and aggressive, even by the standards of lawyers who specialize in representing plaintiffs in high dollar, high-profile cases.

The original complaint alleged that Musk, one of OpenAI’s co-founders — along with Sam Altman, Greg Brockman and others — spun up a strategy to counterbalance what was happening at Google’s DeepMind laboratory where the dark lords were working hard on the development, not just of artificial intelligence, but something far more profound called “artificial general intelligence” or AGI.

There are many different formulations for how to tell when a machine intelligence is properly considered AGI, but a common description is when it can out-think nearly all humans in nearly all human endeavors. The breadth of such an intelligence is the key — that is the “general” in AGI — and it is a clear difference-maker. Regular AI, sometimes called “narrow AI,” can get really, really good at tasks in a single domain — think of autonomous vehicles as an example — but is not good at anything else.

AGI will be good, so it is said, at everything.

According to the complaint, Musk and Altman — both self-described as humanity-centric visionaries — bonded over the fear that Google was on course to create and control AGI. In their view, that meant domination of the human species by the owners of the new technology or worse the actual extinction of humankind by super-intelligent machines. (When Musk tried to talk Larry Page of Google out of the pursuit of AGI, Page reportedly accused Musk of being a “specist.” Musk responded, “Well yes, I am pro-human.”)

Given the epochal stakes, Musk and Altman decided that they could not sit on the sidelines and watch as Google dominated the most powerful AI in history. They would have to throw their chips into the ring of AI development but do it in a way that reflected their lofty values and the sanctity of their mission. They decided to create a new company — OpenAI, Inc. — and charter it as a nonprofit corporation with the mission of pursuing AGI not for crass profit but for the benefit of “humanity” and with the ground rule that the technology it created would be open source, so all the world would have access to the life-changing leaps expected in medicine, science and the arts.

OpenAI was officially formed in December 2015 as a nonprofit corporation in Delaware and received 501(c)(3) tax-exempt status from the Internal Revenue Service on Nov. 3, 2016. That meant charitable donations to OpenAI would entitle the donors to a federal tax deduction, and that meant that the American public, at least indirectly, would be subsidizing the endeavors of the nonprofit. (That tax break would mean, at a high level, of course, that all other taxpayers would have to pay more taxes to cover the tax break the donor obtained.)

Beginning in May 2016 and continuing for four years, Musk donated $38 million to OpenAI and helped recruit world-class engineers to the startup. Altman and Brockman ran the operation and led the on-the-ground strategy.

Google was way ahead, of course, but Altman’s team made rapid advancements. In Aug. 2017, Open AI’s technology defeated the world’s top-ranked human player in one-vs-one Dota 2, a simplified variant of the five-person online battle game where teams work together strategically to destroy the enemy’s structure.

Shaky territory

According to the lawsuit, things began to go awry not long after. On Feb. 21, 2018, Musk resigned from OpenAI’s board of directors (though he continued to provide funding). He had broken away after Altman declined his proposal to combine OpenAI with Tesla and was now developing an AI platform called x.AI that was a full-on competitor to OpenAI.

OpenAI charged forward and began to produce consumer-facing products.

GPT-1 — a research predecessor to today’s ChatGPT — was released in June 2018, and it attracted media attention and early adopter users. While it was a great step forward, it was just the beginning of a far-reaching and mind-bogglingly expensive development effort.

The problem, according to Altman, was that the development costs incurred by OpenAI were vastly beyond what could be supported by a nonprofit corporation, even one bankrolled by one of the world’s richest men. That led to the September 2018 creation of a limited partnership that was a subsidiary of OpenAI (confusingly named OpenAI, LP) and would function on a for-profit basis. The idea was that the for-profit could make commercial arrangements that would induce people who wanted to invest rather than donate to put their money on the table.

To facilitate that possibility, in late 2018 and early 2019, OpenAI transferred substantially all of its intellectual property to OpenAI, LP and received in return a limited partnership interest in the for-profit enterprise. Altman became CEO of the for-profit in April 2019 and did not waste time in taking advantage of the flexibility of the new structure.

Enter Microsoft

With a for-profit subsidiary and the need for staggering quantities of cash, Altman turned to a tech Goliath who did not want to be left out of the potential treasure that could come with AI technology. Microsoft not only had cash, but it controlled lots of “compute” — the tech term for computing capacity — and by July 2, 2019, OpenAI had a billion-dollar investment, and Microsoft had access to OpenAI’s technology and the promise of an ultimate share in OpenAI’s profits.

The structure that would repay Microsoft and other investors was one that Altman would later describe as “bespoke,” a trendy way of saying one of a kind. OpenAI, LP became a “capped profit” enterprise that gave investors a targeted return from future profits, but only up to a fixed amount. For Microsoft, the targeted redemption amount was 20 times its investment (20X caps), that is, $20 billion. Other smaller investors had 100X caps.

According to Altman, Musk was unhappy with these bespoke developments, but he continued his funding. However, there is a dispute over whether he consented to the creation of the subsidiary.

OpenAI raced to show progress. It released GPT-2 in November 2019 and GPT-3 in May 2020. Each model was impressive, but it was the release of ChatGPT in November 2022 that brought the new technology to the attention of the world.

Two months after its launch, ChatGPT had 100 million monthly active users and 13 million unique visitors per day. By January 2023, Microsoft’s investment had swelled to $13 billion.

Ouster and return

But all was not well within OpenAI. On Friday Nov. 17, 2023, OpenAI’s board announced that Altman was no longer CEO nor a member of the board because he “was not consistently candid.” A world class euphemism.

A wild weekend followed — interesting enough to be described at length in Karen Hoa’s 2025 book “Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI,” and dramatic enough to be turned into a lightly fictionalized play called “Doomers” written by Matthew Gasda and performed in both San Francisco and Brooklyn.

On Sunday, Microsoft announced it was hiring Altman and Brockman to start a new AI research group at Microsoft. By Monday, nearly all of OpenAI’s 800 or so employees had signed a letter threatening to leave and go to Microsoft unless Altman was reinstated and the old board resigned.

And then it was over.

Things move fast in Silicon Valley when billions are at stake.

On Tuesday, OpenAI said Altman would return as CEO. Three of the vocal board members would resign, and several new faces would be added, including former Harvard president and former U.S. Secretary of the Treasury Lawrence Summers.

The drama would have been dismissible if OpenAI was selling dental supplies or exercise equipment, but the board of OpenAI was charged with making sure that the vast power of super machine intelligence was devoted to the advancement of “humanity” — not profits.

But now as a result of Microsoft’s power play, the person who was not being “consistently candid” with the board was reinstated, and those board members who had the temerity to expect consistent candor were gone.

That could have been the end of the issue. The vocal old board members had slunk off. Altman had returned triumphantly. The curtain closed on the corporate drama, but against all odds humanity’s champion — a Silicon Valley Spartacus with legal talent on call — decided to get involved.

And so it came to pass that on Feb. 29, 2024, Musk dropped a lawsuit against Sam Altman and OpenAI, Inc. in state court in San Francisco.

The suit asserted several legal theories, but the key one was his contention that, as a nonprofit, OpenAI was formed to benefit humanity and to open source its technology. He argued that OpenAI was not doing that, rather it was enriching its masters, just like any for-profit enterprise, and Musk had been grievously wronged by that turn to the dark side.

Musk’s lawyers spun up an argument that Altman had breached the original “founding agreement” that Musk and Altman had made when they started. They also added the claim that OpenAI was in violation of the corporate laws concerning charitable corporations of Delaware, where OpenAI was chartered, and/or California where it was based.

The suit also had a made-for-media legal theory that Musk explained by referencing a song from the Broadway musical “Annie.”

“OpenAI’s attainment of AGI, like ‘Tomorrow’ in Annie, will always be a day away, ensuring that Microsoft will be licensed to OpenAI’s latest technology and the public will be shut out, precisely the opposite of the Founding Agreement,” his complaint explained.

The suit said that a key component of OpenAI’s deal with Microsoft was the sharing of OpenAI’s technology. Microsoft had the right to use OpenAI’s intellectual property for commercial purposes. However, according to Musk, that right had a kill switch; Microsoft’s right to technology sharing stopped if, as and when, OpenAI achieved AGI. Of course, how that could be determined was a complex issue, but according to Musk, the ultimate decider was OpenAI’s board of directors (the same board that had just been defenestrated when Altman was restored to power and was now refenestrated with supposed Altman/Microsoft approved appointees.)

OpenAI quickly lawyered up. They had San Francisco’s Morrison & Foster on the ground and New York’s legendary Wachtell Lipton in the background. (Wachtell’s William Savitt led the successful suit against Musk filed in July 2022 when he tried to back out of his $44 billion offer to purchase Twitter.)

Musk had a trial superstar from Irell & Manella LLP named Morgan Chu who was described in a legal publication that collects information on lawyers as “beyond doubt the most gifted trial lawyer in the USA” who “delivers staggering results for clients.”

But Chu didn’t get the chance to strut his stuff. OpenAI’s legal filings and coordinated blog posts went hard negative on Musk, painting him as a disgruntled investor animated by a smoldering green-eyed jealousy of OpenAI’s improbable success after he left it for dead.

Altman’s post said, “We’re sad that it’s come to this with someone whom we’ve deeply admired — someone who inspired us to aim higher, then told us we would fail, started a competitor and then sued us when we started making meaningful progress towards OpenAI’s mission without him.”

According to Altman’s version of the story, Musk is the sort of plaintiff who litigates with an ulterior motive — and here that motive was not to do a solid for humanity but to sink a competitor — and the role of any proper judge was to suss that out and quickly, thoroughly, and finally dismiss his case.

OpenAI filed a motion asking the court to do just that, but on June 11, 2024, on the eve of the hearing, Musk voluntarily withdrew the case.

For all it seemed to the public, Altman had won the war without even shooting off his cannons. But that view seriously underestimated Musk’s determination.

In Part Two of this story we describe how Musk returned to litigation two months later, now filing a federal court case before Judge Yvonne Gonzalez Rogers.