Drooid Logo
Back to story perspectives

Full Breakdown

Majority of Americans View AI as Existential Threat, New Poll Finds

By Drooid · · How we work

Public Perception of AI Risk

A recent Politico poll of more than 2,000 U.S. adults reveals that roughly two-thirds of the population believes there is at least a moderate chance that advanced artificial intelligence could destroy humanity. The sentiment cuts across party lines, though it is slightly stronger among supporters of Vice President Kamala Harris than among supporters of former President Donald Trump.

Poll Findings

  • 70 % of Harris voters and 60 % of Trump voters see AI as posing a moderate-or-greater existential risk.
  • 17 % of all respondents are “almost certain” that AI could wipe out humanity, with the same split between the two voter groups.
  • Only 22 % consider AI to pose a very small or no risk.
  • 48 % think AI development should be paused outright; the pause-support rate is 58 % among Harris voters versus 44 % among Trump voters.
  • Conversely, 28 % of Harris voters favor continued development for its benefits, compared with 40 % of Trump voters.
  • 26 % suspect AI leaders’ warnings are primarily a marketing stunt.

The poll was conducted in September and reflects public reaction to recent statements from AI industry figures.

Context: AI Leaders’ Warnings

Shortly thereafter, Anthropic co-founder Dario Amodei published an essay urging the United States’ leading AI firms to cooperate on slowing development. The essay received endorsements from Elon Musk and Sam Altman, CEO of OpenAI, who both emphasized the need for caution.

Potential Policy Implications

The poll’s results suggest a sizable portion of the electorate supports regulatory action, such as pausing or limiting AI research. The partisan differences indicate that any legislative response may face divergent political pressures, with Democrats more inclined toward a pause and Republicans more open to continued development.

Official Responses

Industry leaders have framed their calls for restraint as precautionary rather than alarmist, arguing that coordinated slowing could mitigate unforeseen dangers while preserving innovation pathways. Critics, however, note that a quarter of respondents view these warnings skeptically, interpreting them as strategic messaging rather than genuine concern.