Majority of Americans Feared AI Could Destroy Humanity

A new poll reveals significant public anxiety regarding the potential existential risks of artificial intelligence.

Updated on Sept. 21, 2026 in Artificial Intelligence

Isometric editorial illustration of a solitary, towering server chassis in a dark, geometric hall, representing the systemic scale of AI infrastructure. — A new Politico poll reveals that a majority of Americans believe the rapid development of artificial intelligence poses a potential existential threat to humanity. AI Illustration. Upload story photo >

Live Poll

Do you trust that the current development of artificial intelligence is safe for humanity?

A Politico poll has found that a majority of Americans believe artificial intelligence could eventually destroy humanity. These findings coincide with reports of an Anthropic researcher resigning over concerns that AI could cause human extinction.

Why it matters

The unease stems from the rapid speed of AI development and the potential for self-recursive algorithmic improvement. Industry experts have highlighted a growing fear that advanced systems could pose catastrophic threats to the population.

An Anthropic employee estimated a greater than 10 percent chance that AI could kill humans. Current systems operate within critical computer networks, financial grids, and infrastructure.

The players

Nitesh Chawla

He serves as the Director of the Data, A.I., and Computing Initiative at Notre Dame.

Anthropic

This is an AI research company that has recently seen internal resignations linked to concerns over AI safety.

The details

Researchers are increasingly concerned about the pace of development and how self-improving algorithms interact with existing financial and network systems. Beyond existential threats, current documented risks include the proliferation of deepfakes, cybercrime, discrimination, and privacy violations.

Potential existential threats from AI are projected by some individuals to arrive by the end of the decade.

The Tech Race

The RISE A.I. program at the University of Notre Dame represents the broader academic movement to establish governance for self-improving algorithms. This initiative follows a pattern set by research institutions aiming to balance innovation with the mitigation of systemic technological risks.

While existential threats remain theoretical, the public shift reflects immediate concerns about how AI integration affects daily privacy and security. Users should remain cautious of real-world harms like AI-driven cybercrime, deepfakes, and automated discrimination.

The takeaway

The gap between rapid AI advancement and public trust highlights a critical need for transparent safety frameworks in the tech sector. Readers should prioritize digital literacy to identify AI-generated content like deepfakes while monitoring how these tools influence financial and social systems.