One expert witness, Peter Russell, told jurors he fears an arms race to build artificial general intelligence. Russell, a UC Berkeley computer science professor, was the sole expert called by Elon Musk’s legal team in the case over OpenAI’s corporate structure. He described a tension between the race to reach AGI and efforts to keep systems safe, but the judge limited much of his testimony after objections from OpenAI’s lawyers. That restriction narrowed what jurors heard about long-term risks.

Elon Musk’s attorneys opened the day by putting forward a single expert witness to explain what’s at stake as labs push toward more powerful AI. Peter Russell, a professor at the University of California, Berkeley, has studied AI for decades and signed a March 2023 open letter calling for a six-month pause in frontier model training.

The trial centres on Musk’s claim that OpenAI was founded as a public-interest, safety-focused organisation and then departed from that mission by pursuing commercial ventures. Musk’s lawyers want jurors to see a link between the profit motive and the risks Russell described.

Courtroom testimony

Russell told jurors and Judge Yvonne Gonzalez Rogers that AI development carries many risks. He listed cyber threats, misalignment between model goals and human intent, and the winner-take-all dynamics of labs racing to build AGI. He said those dynamics raise the chance of cutting corners on safety to win first.

OpenAI’s lawyers pushed back. During cross-examination they emphasised that Russell was offering background on AI technology, not an assessment of OpenAI’s corporate governance or its specific safety policies.

They argued his expertise didn't extend to judging how the company organised itself or how it treated safety in practice.

Mastheads

The judge sustained several of OpenAI’s objections. As a result, many of Russell’s broader remarks about existential risk and the global race for AGI didn't make it into the public record at trial. That left jurors with a narrower set of expert evidence focused on technical risks rather than policy or organisational failings.

Expert concerns and context

Russell has long warned about competitive pressure among advanced AI labs. He views the push for AGI as both a technological and social problem. He said, in effect, the fastest lab could gain an overwhelming advantage, which creates incentives to prioritise speed over safety.

His participation in the 2023 pause letter puts him in a cohort of researchers who have publicly urged caution. Musk also signed that letter. Musk later launched his own for-profit AI company, xAI, which shows how those concerns coexist with commercial projects aimed at the same goals.

OpenAI’s attorneys used that contrast to question how persuasive warnings about risk are when founders and investors still move quickly to capture market share. They sought to show the gap between high-level alarm and the practical choices that organisations make while building capability.

Russell’s testimony was meant to establish a basic frame: AI progress brings clear technical hazards. It also brings social stressors because a global race concentrates power and reward in whoever reaches AGI first. Yet the court limited Russell’s ability to link that frame directly to OpenAI’s decisions.

That legal narrowing matters. The trial will turn on whether OpenAI’s shift toward commercial activity breached a founding promise or duty.

To prove that, Musk’s team needs to tie safety rhetoric to corporate decisions. Without expert testimony connecting the two, the jury will hear fewer arguments that safety claims were used as cover for profit-seeking.

Legal teams on both sides appear aware of the high stakes beyond this case. If the jury finds for Musk, it could affect how tech founders structure entities and how investors and boards balance mission and return.

Related Articles

A Musk win could reshape how founders structure companies and investor priorities.

This article was created with AI assistance.