An Anthropic researcher has publicly stated there is a 10% probability that artificial intelligence could eliminate all humans by the end of this decade. Evan Hubinger made the assessment following Jacob Coxon's departure from Anthropic, where Coxon had accused the company and its competitors of "recklessly racing to build superhuman systems they cannot control."
Hubinger acknowledged that Anthropic lacks a concrete plan to ensure the technology remains safe and aligned with human values. He emphasized that AI companies are "locked in a race" to develop increasingly advanced systems "despite the risks." The statement underscores growing concerns within the AI research community about the speed of development and the potential consequences of deploying systems without adequate safety measures in place.