Anthropic researcher puts AI extinction risk above 10%

Anthropic safety researcher Evan Hubinger said he believes AI has more than a 10% chance of killing all humans within the next decade, according to the BBC’s report. The warning arrived as Anthropic withheld its latest model from the UK’s AI Safety Institute; neither Anthropic nor the UK government confirmed the decision. This is a personal assessment, not a measured company forecast, but builders should file it beside the archive’s frontier-containment analysis as a governance signal: independent evaluation and stop authority matter more as models gain autonomy.