Updated

A series of interviews released by nonprofit Palisade Research features about a dozen current and former AI researchers from OpenAI, Google DeepMind, and Anthropic warning about catastrophic risks from advanced AI. Geoffrey Irving, a former OpenAI and DeepMind employee, said he estimates roughly a coin-flip chance of human extinction from superintelligent AI. Adding to the concern, Anthropic's own IPO pitch materials reportedly include a warning that its Claude models could resist shutdown attempts and cause catastrophic harm, an unusual admission from a company marketing itself to investors, underscoring how mainstream AI safety fears have become even among industry leaders.