Sam Altman takes the stand in OpenAI vs. Musk trial as both sides land punches

Sam Altman Takes Stand as OpenAI and Musk Land Competing Narratives

Sam Altman testified in the OpenAI versus Elon Musk trial today, with the case expected to conclude this week. The trial hinges on Musk's allegation that Altman and Greg Brockman manipulated him into funding a nonprofit that they then converted to a for-profit venture, with Microsoft named as an aider in the alleged deception.

The trial has already shifted narratives twice. Early testimony from former OpenAI CTO Mira Murati and board member Helen Toner raised questions about Altman's leadership style and the November 2023 board upheaval. Then Brockman and Siobhan Zillis testified that Musk himself explored scenarios where OpenAI would be folded into Tesla, where Altman might lead Tesla AI efforts, and where Musk could retain control—undercutting Musk's claim to have been defending a pure nonprofit vision. Musk also supported for-profit conversion, they testified, if he could control it.

Yesterday, Microsoft CEO Satya Nadella defended the partnership and testified that Musk never complained to him about any deal violations despite having his contact information. Ilya Sutskever, OpenAI cofounder, testified he spent a year documenting Altman's allegedly manipulative behavior but also said he never promised Musk the company would remain permanently nonprofit. Sutskever's stake in OpenAI is worth approximately $7 billion, a detail that likely shapes how the jury evaluates his motivations.

The core question Altman faced today is whether OpenAI's evolution from nonprofit to for-profit represents a necessary adaptation to build frontier AI or a betrayal of the original mission. Both sides are trying to establish the founders' true intent. OpenAI's questioning established that Altman, like Musk, has been focused on beneficial AI development since at least 2015. A 2015 email from Altman stated: "I think the answer is almost definitely not," regarding whether humanity could be stopped from developing AI. "If it's going to happen anyway, it seems like it would be good for someone other than Google to do it first."

Altman also testified about Musk's past suggestion that he might pass control of OpenAI to his children upon his death—a detail that apparently warranted clarification in court.

On the Tesla scenario, Musk had offered Altman a board seat as part of discussions about folding OpenAI's AI work into Tesla. Altman characterized the offer as both an attempt to address his concerns about having no direction over AI development at Tesla and as a nascent threat: if Altman declined, Musk implied he would pursue AI work independently at Tesla anyway. Mike Isaac, covering the trial for the Times, noted this dynamic mirrors Silicon Valley dealmaking culture—Mark Zuckerberg has made similar overtures to companies he wished to acquire, though typically more explicitly.

Altman criticized Musk's preferred stack-ranking approach to engineer evaluation, arguing that AI research labs need psychological safety to allow researchers to pursue exploratory work—the approach that yielded deep research breakthroughs at OpenAI.

During cross-examination, Musk's lead counsel Stephen Molo pursued a line of questioning designed to paint Altman as fundamentally dishonest, reading through a list of alleged misstatements and asking Altman to confirm or deny each. Altman's demeanor contrasted sharply with Musk's combative stance on the stand; he adopted a more measured tone with attempted humility.

On the core accusation of deception, Altman said: "Clearly, there were misunderstandings and a breakdown of trust. I was not trying to deceive the board. I feel badly for the misunderstandings, but that was never my intent." He also provided emotional context about his brief departure: "I had poured the last years of my life into this. I was watching it about to be destroyed. I was also very angry, hurt, and upset. It felt like an incredible betrayal. It was definitely one of the hardest times of my life."

OpenAI's broader defense strategy has been to depict the board as dysfunctional rather than as victims of deception. Nadella called the directors "amateur hour."

The trial is tracking toward conclusion this week.