A Politico survey asked Americans about their views on the risk of artificial intelligence wiping out humanity. While one-third of respondents saw the risk as “moderate,” 17% assessed it as “almost certain” and 20% as “seriously high.” 70% of those who voted for Kamala Harris in 2024 and 60% of those who voted for Trump saw the risk as moderate. 48% of respondents supported pausing development, while 31% supported continuing it. These figures were 58% among Harris voters and 44% among Trump voters. About half found executives’ warnings sincere, while 26% viewed them as marketing. The “Public First” survey was conducted online among 2 thousand 64 adults on September 13-15. While Dario Amodei, Sam Altman and Elon Musk called for a slowdown, Trump defended innovation.
Why it matters
The survey shows that concern about artificial intelligence is widespread among the US public, but that this concern has not translated into a shared position on completely halting development. Differences in risk assessments and on whether development should continue or be paused indicate that the issue is a debate also shaped by political preferences. This picture also raises the question of how safety warnings from executives of artificial intelligence companies are assessed by the public; the fact that the warnings are not seen as sincere shows that the debate is not limited solely to the existence of risks. The key question that remains open is how the high perception of risk among the public will be reflected in regulation and development policies.