Industry

Evidence So Far in Musk vs Altman OpenAI Trial

The jury trial between Elon Musk and OpenAI leaders Sam Altman and Greg Brockman has revealed emails from 2015 showing the lab's founding discussions and early conflicts. Exhibits include plans for a nonprofit AI effort, Musk's role in shaping its mission, concerns over control, and support like a supercomputer from Nvidia. The case centers on claims that OpenAI abandoned its goal of developing artificial general intelligence for humanity's benefit.

Neura News

Neura News

Neura Market Editorial

April 29, 20266 min read
Evidence So Far in Musk vs Altman OpenAI Trial

Musk v. Altman Trial Reveals Early OpenAI Documents

A federal court in California started the jury trial on Monday in Elon Musk's lawsuit against Sam Altman, Greg Brockman, and Microsoft. The case charges them with breaching OpenAI's charitable trust, fraud, and unjust enrichment. At its heart, Musk argues that OpenAI strayed from its original mission to make artificial general intelligence benefit everyone. Artificial general intelligence refers to AI that matches or exceeds human smarts.

Musk helped start OpenAI with Altman and Brockman. He invested early on. Now he runs xAI, a rival AI company under SpaceX. This marks the newest step in years of legal fights by Musk against OpenAI and its leaders. People who used to work at OpenAI and those near both sides watch closely. A jury decision could shape how OpenAI handles its fast-moving tech and business. Reports say both OpenAI and SpaceX plan to go public this year.

Before the trial, court filings already showed notable messages. Those included emails between Altman and Ilya Sutskever, notes from Brockman's diary, and texts between Meta CEO Mark Zuckerberg and Musk. The trial brings fresh exhibits into view, from OpenAI's first days and even earlier.

Emails trace back to 2015. They reveal OpenAI's roots and initial strains. Key points from public exhibits: Nvidia CEO Jensen Huang supplied OpenAI with a sought-after supercomputer. Musk wrote much of OpenAI's mission statement and shaped its start. Altman pushed to rely on Y Combinator for backing. Brockman and Sutskever fretted over Musk's influence. Musk stressed keeping a nonprofit focused on AI good for all.

Founding Plans in 2015 Emails

Exhibit No. 5: A June 2015 email chain between Altman and Musk. Altman outlined a five-part approach for an AI lab. The goal: "create the first general AI and use it for individual empowerment, ie, the distributed version of the future that seems the safest. More generally, safety should be a first-class requirement." He proposed starting with seven to 10 people in a spare Y Combinator building in Mountain View. For oversight, he listed five: himself, Musk, Bill Gates, Pierre Omidyar, and Dustin Moskovitz. "The technology would be owned by the foundation and used 'for the good of the world', and in cases where it's not obvious how that should be applied the 5 of us would decide," Altman noted. Researchers would get competitive pay, equity in Y Combinator, and upside not tied to their work. He suggested hiring a team leader outside the board.

Altman asked if Musk would join beyond oversight, maybe visiting monthly or backing publicly for hires. He compared it to Peter Thiel's part-time role at Y Combinator. He also mentioned a "regulation letter" for AI rules, offering to keep Musk off it. Musk replied, "Agree on all."

Exhibit No. 7: October 2015 emails. Altman proposed Musk commit $100 million upfront and $30 million more over five years. Bill Gates was not yet in, but Altman hoped soon. He doubted Mark Zuckerberg due to Facebook AI Research. Altman suggested he and Musk form the first Safety Board, adding up to three later as a "'second key' for releasing anything that could be dangerous." Musk said, "Let's discuss governance. This is critical. I don't want to fund something that goes in what turns out to be the wrong direction."

Exhibit No. 12: November 2015 exchange. Musk praised a call with Greg Brockman and called the team strong. He wanted an independent 501(c)(3) nonprofit focused on safe, widespread strong AI. It could earn surplus for reserves. Pay would mix salary and bonuses, convertible to Y Combinator or SpaceX stock. Musk offered Tesla's vast sensor data. He pitched "Freemind" as a name, opposing DeepMind's control. He committed time despite Tesla and SpaceX demands, saying real threats need action. He accepted visible involvement.

Altman suggested sharing Y Combinator space and legal help, names like "Axon" or Turing-linked. Musk liked Turing ideas but not test ties.

The #1 Newsletter in AI

Stay ahead of the AI curve

The most important updates, news, and content — delivered weekly.

No spam. Unsubscribe anytime.

Mission Drafts and Official Start

Exhibit No. 14: December 2015 emails drafted OpenAI's mission and release. Musk aimed to draw talent, noting low pay but right goal and setup. Altman stressed no financial duties let them maximize impact and spread AI.

Exhibit No. 16: OpenAI's articles of incorporation from December 8, 2015. It set OpenAI as a nonprofit for charity. Purpose: "to ensure that artificial general intelligence benefits all of humanity, including by conducting and/or funding artificial intelligence research." It would support safe development, societal study, policy, education, and economics. Tech goes public when apt. No private gain.

Hardware Help and Rising Tensions

Exhibit No. 70: April 2016 email. Musk asked Jensen Huang for an early supercomputer, stressing OpenAI's independence from Tesla as his-funded nonprofit for safe AGI. Huang said, "I will make sure OpenAI gets one of the first ones."

Exhibit No. 388: Photo of Huang delivering it, Musk nearby. Wall quote links to Rickover, echoed in Altman's 2013 post.

Exhibit No. 152: August 2017 email from Shivon Zilis, Musk's chief of staff and later board member. She recapped talks with Brockman and Sutskever. They accepted Musk's time and control varying, but not low time with high control. They sought over $100 million start, citing data center costs. Brockman wanted equal equity. Key worry: Musk's control. Zilis noted, "Is the requirement for absolute control? They wonder if there is a scenario where there could be some sort of creative overrule position if literally everyone else disagreed on direction." Biggest issue: no one-person AGI control forever. "The non-negotiable seems to be an ironclad agreement to not have any one person have absolute control of AGI if it's created... An ironclad 2-3yr minority control agreement." Musk replied, "This is very annoying. Please encourage them to go start a company. I've had enough."

Exhibit No. 153: September 2017 email from Jared Birchall, Musk adviser. Attached cap table: Musk 51.20%, Altman, Sutskever, Brockman each 11.01%, plus employee reserves with names and shares.

The Verge plans to update as more exhibits appear.

Related on Neura Market

More from Neura News