Stuart Russell warns Musk’s lawsuit against OpenAI could fuel an AGI race and weaken safety efforts

UC Berkeley computer science professor Stuart Russell testified in a California courtroom on May 4, 2026, warning that intensifying competition to develop artificial general intelligence (AGI) could create “winner-take-all” incentives that lead AI developers to cut corners on safety.

Russell appeared as an expert witness in a lawsuit filed by Elon Musk against OpenAI. Musk alleges the organization has drisourcesed from its original nonprofit mission.

Safety concerns and “winner-take-all” incentives

Russell told the court that pressure to be first to reach AGI could push companies to deprioritize safeguards, describing the current environment as an “arms race.” He said rapid advances in AI could increase risks ranging from cybersecurity threats to AI misalignment.

Russell also described what he characterized as a fundamental tension between moving quickly toward AGI and implementing adequate safety measures.

Courtroom limits and calls for regulation

Russell said discussion in open court about existential risks from unconstrained AI was curtailed asourceser objections from OpenAI’s attorneys. Even so, he argued that government intervention is needed, including stronger regulation of so-called frontier AI systems.

Russell previously co-signed an open letter in March 2023 calling for a six-month pause in AI research. Musk also signed the letter.

Cross-examination

During cross-examination, OpenAI’s attorneys sought to establish that Russell’s testimony did not directly address OpenAI’s corporate structure or its specific safety policies.