Key Evidence Revealed in Musk vs. Altman Case
The trial between Musk and Altman is in full swing, revealing a trove of exhibits that shed light on the early days of OpenAI. So far, we’ve seen email exchanges, photos, and corporate documents that date back to before the AI lab even had a name. Key highlights include Nvidia CEO Jensen Huang providing OpenAI with a sought-after supercomputer, Musk playing a pivotal role in drafting OpenAI’s mission and shaping its initial structure, and OpenAI CEO Sam Altman expressing a desire to rely on Y Combinator for early backing. Additionally, OpenAI president Greg Brockman and Ilya Sutskever raised concerns about Musk’s level of control, while Musk emphasized the necessity of a nonprofit dedicated to the responsible development of AI for the benefit of all.
Musk’s high-profile lawsuit, which kicked off its jury trial in a California federal courtroom, names Altman, Brockman, and Microsoft as defendants. The suit alleges breaches of the company’s charitable trust, fraud, and unjust enrichment. At its core, the case questions whether OpenAI has strayed from its foundational mission of ensuring that artificial general intelligence—often vaguely defined as AI that matches or exceeds human intelligence—serves humanity as a whole. This lawsuit is just the latest in a series of legal actions Musk has taken against OpenAI and its leaders, stemming from his role as a co-founder and early investor. Musk also heads xAI, a competing AI lab under SpaceX.
Former OpenAI employees and insiders are closely monitoring this lawsuit, as the jury’s decision could significantly influence OpenAI’s operations and its rapidly evolving technology. With both OpenAI and SpaceX reportedly aiming for public offerings this year, the spotlight is brighter than ever.
The discovery phase of the lawsuit has already revealed some eyebrow-raising communications among AI industry leaders, including emails between Altman and Sutskever and diary entries from Brockman. Even text exchanges between Musk and Meta CEO Mark Zuckerberg have come to light. And with the jury trial now underway, even more revelations are on the horizon.
Here’s a detailed rundown of the exhibits made public so far, along with the most significant insights from each. While not every item is particularly riveting, we’ve marked the most crucial ones with an asterisk. The list will be updated as new exhibits emerge.
*Exhibit No. 5*
A June 2015 email exchange between Altman and Musk outlines a five-part plan for an AI lab aimed at “creating the first general AI and using it for individual empowerment.” Altman proposes starting with a small team and expanding, suggesting governance by a group that includes himself, Musk, Bill Gates, Pierre Omidyar, and Dustin Moskovitz. He emphasizes that the technology should be owned by the foundation and used for the greater good, with decisions made collectively by the five. Altman also discusses compensation for researchers and the need for a team leader who wouldn’t be on the governance board. He asks Musk if he would be involved in the lab beyond governance, potentially visiting monthly to discuss progress. Musk agrees to all points.
*Exhibit No. 7*
In an October 2015 email, Altman proposes a $100 million commitment from Musk, along with an additional $30 million over five years. He mentions that while Gates isn’t yet committed, he hopes to secure his support soon, and expresses doubts about Zuckerberg’s involvement due to his own AI initiatives. Altman suggests they both serve on a Safety Board, which would be crucial for overseeing any potentially dangerous developments. Musk stresses the importance of governance, indicating he doesn’t want to fund a project that could veer off course.
*Exhibit No. 12*
In a November 2015 email, Musk shares his enthusiasm for the team and suggests establishing the lab as an independent nonprofit focused on the positive development of strong AI. He discusses employee compensation, proposing a cash salary with options to convert to stock in Y Combinator or SpaceX. Musk also offers extensive real-world data from Tesla for the lab’s use and suggests the name “Freemind” to reflect the goal of creating accessible digital intelligence. He acknowledges the need for his involvement, despite potential conflicts with his commitments to SpaceX and Tesla.
*Exhibit No. 14*
A December 2015 email exchange drafts the mission statement and press release for OpenAI. Musk emphasizes that the release aims to attract top talent, while Altman focuses on the potential for positive human impact without financial obligations.
*Exhibit No. 16*
OpenAI’s official articles of incorporation, filed in December 2015, declare it a nonprofit dedicated to ensuring that artificial general intelligence benefits all of humanity. The document outlines the corporation’s commitment to conducting and funding AI research while analyzing its societal impacts.
*Exhibit No. 70*
An April 2016 email from Musk to Nvidia CEO Jensen Huang requests an early supercomputer unit for OpenAI, clarifying that it is a nonprofit initiative. Huang assures Musk that OpenAI will receive one of the first units.
*Exhibit No. 388*
A photo captures Jensen Huang delivering the supercomputer, with Musk present. A quote attributed to Admiral Hyman G. Rickover is visible in the background, echoing sentiments shared by Altman in a previous blog post.
*Exhibit No. 152*
In an August 2017 email, Musk’s chief of staff, Shivon Zilis, summarizes a meeting with Brockman and Sutskever, highlighting concerns about Musk’s control over the company. Zilis notes that while they are open to Musk having less control if he spends less time on the project, they are wary of any scenario that grants him absolute control over AGI. Musk’s response indicates his frustration with the situation.
*Exhibit No. 153*
A September 2017 email from Musk’s adviser, Jared Birchall, presents a simplified cap table proposal, showing Musk with a 51.20% equity stake, while Altman, Sutskever, and Brockman each hold 11.01%. The document also includes reserved equity for employees.
