OpenAI board member warns advanced AI could be 'deadly' without safety measures

Paul Christiano, a top US official appointed to OpenAI's non-profit board, publicly warned that advanced AI could 'kill most people' absent stronger safety features. The statement signals internal concern over AI risk from a government-linked figure, though it remains an individual opinion rather than a confirmed policy shift. This matters as it could influence AI regulation and public perception of frontier AI development.