AI Governance2026-11-15The Verge

Helen Toner says AI CEOs still can’t be trusted

Helen Toner, a former OpenAI board member who was involved in Sam Altman's brief ouster, argues that the public still cannot fully trust AI chief executives. In an interview with The Verge, Toner, now executive director of Georgetown's Center for Security and Emerging Technology, discussed AI safety, governance, and the concentration of power in a small number of companies. Her central point is not that AI leaders are uniquely dishonest, but that the public lacks the independent oversight needed to verify their claims and hold them accountable. That argument lands at a moment when AI companies are asking for trust on issues ranging from model risks to deployment decisions, even as key information remains under corporate control. Toner's background gives the critique extra weight: she was part of the OpenAI board during the dramatic 2023 dispute over Altman's leadership, an episode that exposed deep questions about governance at the most prominent AI lab. The interview is part of a broader conversation about AI in real life, but its focus on corporate accountability is especially timely. As AI systems become more capable and more embedded in daily life, the debate is shifting from whether executives promise safety to whether external institutions can check those promises. Toner's view suggests that trust should be earned through transparency, independent evaluation, and clear consequences, not through personal reputation alone. For policymakers and the public, the lesson is to treat AI leadership claims with scrutiny and to build governance that does not depend on trusting any single CEO.

Related news