Stuart Russell, a computer science professor at the University of California, Berkeley, testified as Elon Musk's sole AI expert witness in the ongoing trial between Musk and OpenAI. Russell presented concerns regarding the risks associated with the pursuit of artificial general intelligence (AGI), particularly the potential for an "arms race" among leading AI laboratories. He described this competition as creating a fundamental "tension between the pursuit of AGI and safety."
During his testimony, Russell explained that the immense economic value associated with being the first to achieve AGI creates powerful financial incentives. This "winner-take-all" dynamic, he warned, could lead to a situation where a small number of companies control a significant portion of global economic activity. These substantial economic stakes, Russell suggested, may cause safety considerations to become a secondary concern for developers.
Russell characterized the safety of AI systems as an "unsolved scientific problem." He argued that until robust safety protocols are established and validated, companies will feel compelled to continue their development efforts to avoid falling behind. This competitive pressure, he stated, drives the ongoing race to create more powerful AI systems.
The professor's testimony provided jurors with a framework for understanding the competitive pressures surrounding AGI development. However, his role was not to scrutinize OpenAI's corporate documents or assess its internal safety policies. Instead, Russell focused on articulating the broader pressures and incentives that encourage rapid AGI development, even at the potential cost of caution.
Russell's background lent credibility to his warnings. He previously signed an open letter in March 2023 advocating for a six-month pause in advanced AI development to establish safety protocols. He has also consistently called for increased oversight of frontier AI laboratories.
The court, presided over by U.S. District Judge Yvonne Gonzalez Rogers, limited the scope of Russell's testimony. Discussions concerning hypothetical long-term existential risks posed by AI were largely excluded, as the judge indicated these were not central to the legal questions at hand.
Russell's compensation for his expert report was revealed to be $5,000 per hour for initial preparation, followed by $1,500 per hour thereafter, with his total earnings from the case projected to reach approximately $235,000. These payments were made through Excession LLC, Musk's family office. This rate is notably higher than the typical $500 to $1,000 per hour charged by expert witnesses in high-profile legal disputes.
Concerns about an AI "arms race" are not new for Russell. In July 2025, he expressed similar sentiments at the World Artificial Intelligence Conference in Shanghai, urging developers to prioritize safety over the race to create advanced, opaque AI technologies. He noted that the current scale of investment in AI is immense, and a failure to yield quick returns could lead to an industry bubble.
Russell also highlighted societal issues stemming from rapid AI deployment, including potential job losses, surveillance concerns, and misuse of the technology. He pointed to growing public unease, particularly among younger generations, regarding the dehumanizing aspects of AI.
The trial continues to examine Musk's lawsuit, which alleges that OpenAI betrayed its original charitable mission by prioritizing profit and a for-profit structure. The proceedings aim to determine the legality of OpenAI's corporate restructuring and its implications for the future of AI development and governance.