New Survey Reveals Majority of Americans Favor Slowing or Halting AI Progress

A Quinnipiac University poll has put measurable public opinion behind a question growing louder around AI development: how fast should the technology move? The survey found that 47 percent of US adults want to reduce the pace, while 30 percent want a full stop until safety has been verified. Only 5 percent want faster development. The result matters because it moves the conversation from what AI companies can build to what the public will accept, and from company assurances to independent safety standards.

Background: Safety Questions Before the Poll

The poll arrived as public debate about AI safety has focused on whether oversight should come from outside the companies building these systems. The central tension is familiar: faster releases can spread useful tools more quickly, but they can also outpace checks on reliability, misuse, and long-term risk. The survey does not measure technical progress itself. It measures how Americans view that progress after a series of public warnings and debates about the direction of advanced AI.

Independent safety standards have become one possible answer to that tension. The Quinnipiac University poll found 86 percent support the independent safety standards currently under debate for AI companies, even if they slow progress. That number is important because it links two issues often treated separately: public concern about risk and public willingness to accept slower deployment if oversight appears stronger.

Development: The Numbers That Define the Poll

The Quinnipiac University poll surveyed 1,202 US adults from September 24 to 27, with a margin of error of +/- 3.5 percentage points. The sample gives a snapshot of national attitudes during the late September survey window, not a census of every community.

The clearest split is between speed and caution. Forty-seven percent want to reduce the pace. Thirty percent want a full stop until safety has been verified. Together, those responses point to broad support for slowing AI development, even though the preferred method differs. One group wants a gentler curve; the other wants a pause.

The risk questions deepen the picture. Seventy-three percent worry that AI could eventually threaten humanity's survival. That is not a small-share concern. It suggests that a majority of respondents see the technology as capable of producing consequences beyond ordinary product failures or everyday annoyances.

The trust numbers help explain why caution is rising. Seventy-four percent have little or no trust in the leaders of AI companies. Public confidence does not appear to rest mainly on reassurance from executives. Instead, the poll shows demand for verification outside the companies themselves.

Chetan Jaiswal of Quinnipiac University summarized the pattern: "Taken together, these results suggest that the public wants safeguards that do not depend simply on taking company leaders at their word."

Latest: A Trust Gap Behind the Numbers

The current stage of the debate can be read as a collision between innovation timelines and public confidence. The Quinnipiac University poll does not show that most Americans oppose AI development. It shows that most want a different pace. The numbers separate the desire for useful technology from trust in the institutions delivering it.

That distinction matters for AI companies and for policymakers. A company may argue that its systems are safe enough for release, but if 74 percent of respondents have little or no trust in the leaders of AI companies, that argument starts from a weak public foundation. The 86 percent support for independent safety standards, even if they slow progress, suggests that external checks are not seen as bureaucratic friction alone. They are being viewed as a way to make the pace acceptable.

The margin of error, +/- 3.5 percentage points, means the broad direction is clear even if exact percentages could shift slightly. The survey window from September 24 to 27 places these views in a short period of public discussion, not in a record of model releases or formal policy decisions.

What to watch next is whether the debate over independent safety standards moves from polling support to concrete requirements that AI companies must meet. The Quinnipiac University poll shows a public willing to trade speed for verification, and it shows a trust gap that may shape the next stage of AI development policy.

Meta description: Quinnipiac University poll: 47 percent want AI development slowed, 30 percent want a full stop, and 86 percent back independent safety standards.

Tags: Quinnipiac University poll, AI safety, AI development, public opinion, independent safety standards

Featured image: Abstract technology style graphic with a timeline, shield, and percentage bars; no real people, no company logos, no text