New Musk lawsuit against OpenAI claims Altman ‘betrayed’ him

Elon Musk has decided he wants to bring Sam Altman and OpenAI to court after all in a brand-new lawsuit over whether OpenAI is actually open and not-for-profit.

The complaint [PDF], filed today in a California federal court, is largely a redo of his suit from February this year, which was silently dropped in June. In fact, the two complaints share much of the same text, though the new complaint is about twice as long and charges Altman and OpenAI with 15 counts of wrongdoing instead of the initial five. Musk has also seemingly hired a new legal team.

In short, the Tesla tycoon claims he was fraudulently tricked into supporting the launch and growth of the ChatGPT maker, that what the lab became is not what he was sold from the start, and as such he wants payback for being allegedly conned.

'Greed'

In the latest filing, the Musk team added a preamble to the complaint. "Elon Musk's case against Sam Altman and OpenAI is a textbook tale of altruism versus greed," the complaint alleged. "Altman, in concert with other defendants, intentionally courted and deceived Musk, preying on Musk's humanitarian concern about the existential dangers posed by artificial intelligence."

The complaint accuses Altman and his friends of making promises they knew they wouldn't keep when they founded OpenAI with the backing of Musk all the way back in 2015. For Musk, it was important that the lab's research was openly available, if not open source, and safe; OpenAI's articles of incorporation seemingly certifies that its mission is to make its technology to "benefit the public."

There were apparently lots of perks for OpenAI with Musk as an ally, and the filing argues that his involvement gave the organization "connections, credibility, and clout." The SpaceX supremo also allegedly sent OpenAI $44.6 million between 2016 and 2020, assuming that it would be going toward initiatives that Musk believed in.

But according to the complaint, things started to go wrong when Altman and OpenAI co-founder Greg Brockman (who is also a defendant) discussed making OpenAI a for-profit firm in 2017 and 2018. Musk apparently strongly disapproved of the idea, but OpenAI launched for-profit parts of its business the very next year, which the suit says go so far as to be a "scheme to control and cash in on OpenAI, Inc's technology."

When OpenAI declined to make its AI models from GPT-4 onward openly available, that apparently was when Altman "betrayed" Musk.

The filing variously calls out OpenAI's business dealings with Microsoft, Reddit, Helion Energy, and other companies affiliated with Altman; it also doesn't neglect to point out the copyright lawsuits OpenAI is defending against, as well as inquiries from the FTC, SEC, and the US Senate. The Musk legal team seems to have thrown everything but the kitchen sink at the chatbot outfit.

Don't forget that, post-OpenAI, Musk set up his own generative AI lab called xAI, whose chatbot assistant Grok is baked into his social network Twitter. We're in no doubt Elon would love to inject OpenAI's technologies into xAI and Tesla – especially after funneling millions of dollars into OpenAI – but can't or won't because OpenAI doesn't share its toys for free.

Potentially succeeding where the old suit failed

The latest suit is essentially the same lawsuit as before but on steroids.

The first attempt alleged breach of contract, citing the articles of incorporation of OpenAI as evidence that what was promised – all that grand talk of openness – was not delivered. The new one goes further, presumably in hope that the more it lobs at OpenAI, the more that will stick.

Count VI alleges a "breach of express contract," the kind of contract made explicit through some sort of written agreement. Here, the Musk team says 2015-era correspondence between the X boss and Altman constituted a contract, wherein Musk agreed to supply capital and OpenAI would produce open technology in a non-profit capacity. Musk reckons that promise by OpenAI was not met, which would be a contract breach.

Count VII covers a "breach of implied-in-fact contract," which means a contract formed through actions rather than explicit words. The reasoning here is similar, and brings up the articles of incorporation again.

On a similar note, count VIII is for "breach of implied covenant of good faith and fair dealing," and count IX is for "breach of quasi-contract/unjust enrichment." Clearly, the Musk legal team is trying to cover all bases and establish that there was a breach of contract at some point, which perhaps the original complaint failed to do sufficiently.

• Musk ropes Dell, Supermicro into xAI supercomputer project
• Tesla shareholders agree to pay Musk staggering sum of $48B
• Musk wants to ban Apple at his companies for cosying up to OpenAI
• Elon Musk confirms 12K H100s ordered for Tesla were instead prioritized for xAI

The other counts cover allegations of fraud, racketeering, false advertising, unfair competition, breach of fiduciary duty, and tortious interference. Ironically, the unfair competition charge and one of the false advertising charges relies on local law in California, a state that Musk is trying to get X and SpaceX out of. (Just this week, Twitter said it's storming off from its San Francisco office to use space in nearby San Jose and Palo Alto.)

According to the complaint, Musk wants much more than just cash, since the damage OpenAI has allegedly done "cannot be reasonably or adequately measured or compensated in money damages." The big ticket items are a "constructive trust" of everything OpenAI gained thanks to Musk's donations, plus the end of the licensing deal that gives Microsoft exclusive access to GPT-4-based models.

Which would suit xAI and Tesla.

But Musk is also seemingly asking for lots of money, including all of the profit OpenAI made with the $44.6 million donations, plus treble damages, punitive damages, and legal fees.

The Register reached out to OpenAI for comment, and it reiterated what it said last time Musk sued, saying "Elon's prior emails continue to speak for themselves." OpenAI published the emails in question in March. ®