Anthropic researchers warn of non-zero AI extinction risk, mainstreaming doomsday fears

Anthropic researchers publicly acknowledged a non-zero chance of AI causing human extinction within a decade, with one researcher stating no other human activity poses this level of danger. This marks a shift from fringe concern to mainstream discourse as AI capabilities grow. The claim is speculative and based on expert opinion, not empirical evidence, but its public endorsement by a leading AI lab signals growing institutional concern.