As Elon Musk moves to dismiss a lawsuit from Tesla shareholders claiming he failed his fiduciary duties, OpenAI is now suing Musk and exposing lies, which could help Tesla shareholders.

It looks like cracks are forming in Elon Musk’s armor of lies.

Last year, Tesla investors sued Musk for breach of fiduciary duty and resource tunneling over the founding of xAI, a private AI company under his control.

Musk had previously stated that Tesla would be a major player in AI and that AI products would be critical to Tesla’s future, but in early 2024, the CEO threatened not to build AI products at Tesla if he didn’t get more control over the company by getting more shares.

He also redirected NVIDIA AI hardware acquired by Tesla to xAI and hired Tesla employees to work at xAI.

In short, Tesla shareholders argue that Musk is in breach of his fiduciary duties to shareholders by creating a private company that competes directly with Tesla. The lawsuit also cites similar issues with Musk’s acquisition of Twitter.

This week, Musk and Tesla board members, who are also defendants in the lawsuit for not stopping Musk, have filed to try to dismiss the lawsuit (via Bloomberg):

The shareholder suit by a group of pension funds and other investors “is long on hyperbole but woefully short on well-pled facts,” the board members said in a court filing Monday. “Yet they cannot escape the undeniable reality: Tesla has thrived under this board and CEO, delivering astronomical returns to stockholders while advancing its mission to create sustainable abundance for all.”

It will likely take a while before the lawsuit moves through the court, but in the meantime, Tesla shareholders have found a strong ally: OpenAI.

Musk has long been tormenting OpenAI with lawsuits. Tesla’s CEO co-founded OpenAI as a non-profit in 2015 to develop an artificial general intelligence that positively contributes to humanity.

In early 2018, Musk resigned from OpenAI, citing “conflicts of interest with Tesla.”

At that time, Tesla’s CEO started pushing the automaker increasingly toward self-driving, which he often described as “real-world AI,” and the automaker began to compete for AI talent with OpenAI.

While he was seemingly on good terms with OpenAI after his departure, a few years later, he started publicly criticizing the organization for moving to a limited for-profit model, which they argued was necessary due to the billions of dollars required to build the compute training hardware to have an impact in the AI sector.

Musk even sued the company over the move and repeatedly publicly mocked them.

It hasn’t been clear how serious the legal actions have been since Musk even claimed that he would drop the lawsuit if OpenAI changed its name:

OpenAI has been defending itself with the release of some emails that show Musk actually agreed for years that the organization needed to move to a for-profit model.

Now, OpenAI has countersued Musk and released more details that show Musk has been misleading the public for years.

It also explains his latest moves at Tesla and xAI.

Musk’s AI effort went from OpenAI to Tesla to xAI

All the documents released by OpenAI as part of the countersuit paint a much clearer picture of Musk’s involvement with AI and how it evolved over the years.

I’ll start with a clear timeline to make it easier to understand.

• 2010s: Musk has long been fascinated with AI and emerged as one of the most prominent tech voices warning about its dangers.
• 2015: Musk co-founds OpenAI as a non-profit to try to create a safe AGI.
• 2017: Musk privately communicates to many people in OpenAI and the AI community that the company needs to switch to a for-profit model and raise billions to be successful due to the cost of AI hardware.
• 2018: Musk attempted to get control of OpenAI and merge it with Tesla, but this was rejected by OpenAI’s board, which ultimately took investments from Microsoft to start its for-profit arm, as it gave the organization more independence.
• 2018: Musk leaves OpenAI, citing a conflict of interest with Tesla.
• 2018-2022: Musk positions Tesla as “the world’s leader in AI”, hires a ton of AI talent, and claims Tesla will “play an important role in AGI”.
• 2022: Musk sells tens of billions of dollars worth of Tesla stocks, partly to buy an overpriced Twitter.
• 2023: Shortly after the viral launch of OpenAI’s ChatGPT, Musk creates a new private company, xAI, to develop AI products and compete with OpenAI.
• 2024: A judge rescinds Musk’s $55 billion Tesla CEO compensation package, which would have increased his stake in Tesla back to where it was before he bought Twitter.
• 2024: Musk threatens Tesla shareholders that he will not build AI products at Tesla unless he gets more control (aka more shares).
• 2024: Musk hires Tesla employees for xAI and redirects shipments of AI training compute from Tesla to xAI.
• 2024: Musk sues OpenAI to try to block its transition into a capped for-profit business.

OpenAI has all the receipts to prove this. I recommend reading all the emails because they give great insights into Musk’s persona and how he presents himself publicly versus what he says privately.

Here are some of the highlights to prove the timeline above:

Early on in the founding of OpenAI in 2015, it was proposed to be a non-profit linked to Y Combinator, Sam Altman’s company at the time, and Musk was already suggesting to make it a regular C corp:

OpenAI shared many internal emails and text messages between the teams, Musk, and Musk’s executive assistant/future baby mama, Shivon Zilis, discussing the need for much more capital, which will require a move to for-profit.

In 2017, as OpenAI was first configuring a potential for-profit arm, Musk tried to take control by asking for preferred shares and a supermajority:

Musk even filed for a new benefit corporation, a for-profit legal structure that aims to generate profits while positively impacting society and/or the environment.

Musk’s full-time money manager, Jared Birchall, is listed as the sole director of the new corporation.

OpenAI rejected Musk’s proposal as it would have given him complete control, but they insisted they still wanted to work with him.

In early 2018, Musk switched up his proposal to try to get OpenAI attached to Tesla:

This proposal also failed, as OpenAI felt this was also an attempt from Musk to gain complete control.

Musk then left OpenAI and focused his AI efforts on Tesla until he significantly reduced his stake in the company to buy Twitter on a whim.

Then, he founded xAI to become his main AI effort as a private company under his control while telling Tesla shareholders that the company was an “AI and robotics play.”

xAI recently absorbed X (Twitter), resulting in a $125 billion company based on Musk’s made-up valuation.

Electrek’s Take

This is extremely revealing. It clearly shows that Musk’s main goal is to have complete control over AI.

He tried to get control of OpenAI, but couldn’t make it work. He then tried to make it work with Tesla, but he screwed up by giving up some control (I’d argue he still has a firm hold on the public company) through the acquisition of Twitter.

He panicked after OpenAI launched ChatGPT and started xAI as a private company entirely under his control, devaluing Tesla in the process – hence the current shareholders’ lawsuit.

For years, Musk attacked OpenAI and lied to the public about disagreeing with the for-profit transition, when he was actually pushing for it since the very beginning. The only difference is that OpenAI was now a competitor to Tesla, and then xAI.

I want to be clear here. OpenAI is not completely clean, either. It obviously owes Musk something for the company’s original funding, but the emails also reveal that the organization tried to give him shares and pay him back, but Musk refused.

His refusal is likely linked to his believing that he could do more damage by suing OpenAI.

It looks like Musk believes that he is some sort of super genius who deserves to be the one in control of a potential future AGI, and he was willing to lie and cheat his way into making it happen.

Even if his intentions are good, that’s a scary thought.