A federal court in Oakland, California, recently heard testimony regarding Elon Musk’s lawsuit against OpenAI, which centers on allegations that the frontier AI lab has strayed from its foundational mission of ensuring artificial general intelligence (AGI) benefits humanity. The legal proceedings are examining whether the company’s for-profit subsidiary has compromised its original non-profit ethos and commitment to AI safety.

Key testimony came from Rosie Campbell, a former employee who joined OpenAI’s AGI readiness team in 2021 and departed in 2024 after her team was disbanded, alongside the Super Alignment team. Campbell highlighted a significant shift within the organization, stating that it transitioned from a “very research-focused” environment where AGI and safety issues were openly discussed, to a “product-focused organization.” She expressed concerns that developing a super-intelligent AI without robust safety protocols would contradict the company’s initial mission. Campbell cited an incident where Microsoft deployed a version of OpenAI’s GPT-4 model in India via its Bing search engine before the company’s internal Deployment Safety Board had evaluated it. While acknowledging the necessity of substantial funding for AGI development, she stressed the importance of establishing strong safety precedents as AI technology advances, even if the immediate risk was not substantial. Under cross-examination, Campbell conceded that, in her view, OpenAI’s safety approach surpasses that of xAI, the AI enterprise founded by Musk.

Further insights into OpenAI’s internal governance were provided by Tasha McCauley, a former non-profit board member. McCauley testified about concerns regarding CEO Sam Altman’s transparency and management style, which she described as “conflict-averse.” She detailed instances where Altman allegedly misled the board, including failing to inform them about ChatGPT’s public launch and not disclosing potential conflicts of interest. These governance issues, she explained, ultimately led to the non-profit board’s brief decision to remove Altman in 2023. The board’s ability to oversee the for-profit entity was severely compromised by a lack of trust and insufficient information, a situation that was only reversed after employees largely sided with Altman and Microsoft intervened.

This apparent failure of the non-profit board to effectively influence the for-profit organization directly supports Musk’s central argument that OpenAI’s transformation into a major private entity has fundamentally broken its implicit founding agreement. David Schizer, a legal expert testifying for Musk’s team, reiterated the importance of prioritizing safety over profits, emphasizing that taking safety rules and processes seriously is paramount. The ongoing legal battle underscores a broader debate about the governance of advanced AI, with McCauley suggesting that internal failures at OpenAI highlight the need for stronger governmental regulation of the burgeoning technology, particularly when the public good is at stake.