9/30/26

Following recent warnings about rogue AI agents, sparking debate over safety versus economic advantage as the U.S. and China race to develop AI systems, 73 percent of Americans are either very concerned (38 percent) or somewhat concerned (35 percent) that future AI systems could potentially threaten human survival, while 25 percent are either not so concerned (14 percent) or not concerned at all (11 percent). This is according to a Quinnipiac (KWIN-uh-pea-ack) University national poll of adults released today examining attitudes about artificial intelligence. The survey was conducted in collaboration with the Quinnipiac University School of Computing & Engineering and the Quinnipiac University School of Business.

Fifty-three percent of Americans think AI will do more harm than good in their day-to-day lives, while 34 percent think AI will do more good than harm in their day-to-day lives, with 13 percent not offering an opinion. This is little changed from Quinnipiac University's Poll on AI from March 30, 2026.

When asked which concerns them more, over half of Americans (52 percent) say they are more concerned about AI being used by humans to do harm to humanity, while 33 percent say they are more concerned about autonomous AI agents doing harm to humanity and 2 percent volunteer no harm to humanity, with 12 percent not offering an opinion.

A vast majority of Americans say they do not have trust in the leaders of AI companies.

Seventy-four percent say they have either not so much trust (29 percent) or none at all (45 percent) in the leaders of AI companies, while 24 percent say they have either a lot of (3 percent) or some (21 percent) trust in the leaders of AI companies.

Ninety-one percent of Americans think it is either very important (74 percent) or somewhat important (17 percent) that the United States establish guardrails for AI systems, while 7 percent think it is either not so important (3 percent) or not important at all (4 percent).

A vast majority of Americans (86 percent) support requiring companies that are developing AI systems to meet independent safety standards, even if it slows down AI development, while 9 percent oppose it.

Americans were given four options about the pace of AI development and asked which one comes closest to their line of thinking about what AI companies in the U.S. should do:

More than two-thirds of Americans (69 percent) think it is either very important (40 percent) or somewhat important (29 percent) that the United States keep up with China in developing AI systems, while 27 percent think it is either not so important (14 percent) or not important at all (13 percent).

Eighty-one percent of Americans think safety is more important than staying at the forefront of innovation when it comes to AI, while 14 percent think staying at the forefront of innovation is more important than safety.

Seventy-two percent of Americans say they would oppose the building of an AI data center in their community, while 21 percent say they would support it, with 7 percent not offering an opinion.

In Quinnipiac University's Poll on AI in March, 65 percent said they would oppose the building of an AI data center in their community, 24 percent said they would support it, and 11 percent did not offer an opinion.

1,202 U.S. adults nationwide were surveyed from September 24th - 27th with a margin of error of +/- 3.5 percentage points, including the design effect.

The Quinnipiac University Poll, directed by Doug Schwartz, Ph.D. since 1994, conducts independent, non-partisan national and state polls on politics and issues. Surveys adhere to industry best practices and are based on probability-based samples using random digit dialing with live interviewers calling landlines and cell phones.

Visit poll.qu.edu or www.facebook.com/quinnipiacpoll

Email poll@qu.edu or follow us on X @QuinnipiacPoll.