Politico Poll: 63% of Americans View AI as Existential Threat, Plurality Favor Development Pause
Summary
- • 63% of Americans say AI poses at least a moderate risk of destroying humanity
- • Plurality favor pausing advanced AI development over continued expansion
- • Poll follows rare alignment: Amodei, Altman, and Musk all called for AI slowdown
- • Trump dismissed AI safety warnings as a 'HOAX' on Truth Social
Details
Survey conducted Sept 13–15, 2026
Independent poll by London-based Public First for POLITICO, fielded three days after AI lab leaders publicly called for slowing development.
63% see AI as existential risk
About two-thirds of Americans say there is at least a 'moderate' risk that advanced AI could destroy humanity — a majority across parties, genders, and age groups.
Plurality favor pausing AI development
More Americans lean toward slowing or pausing advanced AI model development to reduce risk than toward continuing to build more powerful systems.
Partisan gap: 70% Harris vs. 60% Trump voters
Harris voters are somewhat more likely to see existential AI risk (70%) than Trump voters (60%), though both groups form majorities.
AI lab leaders united in slowdown call
Anthropic's Dario Amodei and OpenAI's Sam Altman joined Elon Musk in a rare show of alignment, publicly calling for slowing advanced AI development.
Trump dismisses safety alarms as 'HOAX'
Despite the industry warnings and public concern, Trump posted on Truth Social dismissing AI safety alarms and insisting AI development will not be stopped during his term.
Democrats weighing House select committee on AI
Some Democrats are considering creating a House select committee on AI if they win back the House in November; others have called for a ban on artificial superintelligence.
Public trust in AI leaders' warnings is high
Roughly half of U.S. adults believe AI leaders warning of existential risk 'will kill us all' are doing so genuinely; only 26% view it as a marketing tactic.
Americans accept geopolitical tradeoff on safety
Respondents split in favor of more responsible AI development even if it means falling behind China — a significant finding given the dominant 'AI race' framing.
Broad discomfort with AI in daily life
Across a range of everyday scenarios, majorities of Americans said they would be uncomfortable with AI being used in their daily lives.
Poll data from The POLITICO Poll conducted by Public First, Sept 13–15, 2026.
What This Means
A major Politico poll reveals a striking hardening of American public opinion: nearly two-thirds now view AI as a genuine existential threat, with bipartisan majorities holding that view. This comes as top AI lab leaders took the rare step of jointly calling for caution, injecting long-simmering safety concerns into mainstream political debate. The divergence between growing public fear and the current administration's pro-acceleration stance positions AI safety as a significant political fault line heading into the November elections. Congress is beginning to respond, with proposals ranging from select committees to outright bans on artificial superintelligence.