Nature weighs AI extinction warnings against the science

Independent AI-safety researchers told Nature that the extinction warnings now coming out of frontier labs rest on thin scientific ground. The warnings came back into focus in September, when Anthropic researcher Jacob Coxon resigned, saying people building AI "earnestly believe that it could kill us all by the end of the decade." Anthropic's alignment-science lead Evan Hubinger has publicly put the risk of human extinction at ">10% within the next decade."

Michael Vermeer, a technology-policy researcher at RAND, told Nature that such forecasts "involve so many untestable claims that you just end up with a conversation that is really more like faith than something scientific or empirical." A 2025 RAND study he worked on concluded that complete extinction via nuclear weapons is not feasible. Biotechnology and atmospheric-modification routes were not ruled out, but any AI attempting them would need substantial physical capabilities, and the attempts would likely be detectable in time for humans to intervene.

Heidy Khlaaf, chief AI scientist at the AI Now Institute in New York City, calls the extinction framing "fear-mongering," arguing that disinformation, psychological harm, bioweapon creation and military miscalculation deserve the attention instead. Nature cites a CNN report of AI-generated false intelligence that nearly caused a U.S. military incident with a Chinese vessel.

The piece was moving through analyst circles on our radar quickly; four of the AI experts we track posted the Nature link.