Most Americans think artificial intelligence is moving too fast, and it’s one of the few things Democrats and Republicans agree on. In a new AP-NORC poll, 64% of adults said AI is developing too quickly in the U.S. Only 8% said it’s moving too slowly.
Even a majority of Republicans, 54%, said it’s going too fast, along with 76% of Democrats, according to AP-NORC.
What do Americans want the government to do about AI?
About 8 in 10 said the government should make keeping AI under human control and protecting U.S. workers top priorities. Democrats and Republicans gave similar answers on those, and on promoting innovation. They split more on whether the government should focus on limiting AI’s environmental harm and on competing with other countries.
Who do Americans trust to handle AI?
Not many people. Just 31% approve of how President Trump is handling AI. Twenty-three percent trust Democrats more on the issue and 21% trust Republicans more, but the biggest group, 39%, said neither party can be trusted to do a good job.
Are people using AI to decide how to vote?
Not much. Ahead of the November 3 midterms, 21% said they’ve used or plan to use AI to learn about a candidate or ballot issue. Even fewer said they’d use it to find out how to vote (12%), how to register (10%) or who to vote for (10%).
The poll surveyed 2,140 adults September 24-28 and has a margin of error of plus or minus 2.9 points.
The BeezLoop Take
When nearly two-thirds of the country agrees on anything, Washington should pay attention. People aren’t saying stop AI. They’re saying slow down and keep humans in charge, and protect the workers who’ll feel the changes first. That’s a reasonable ask.
The industry keeps proving the point. This month an AI model sent a fake murder tip to Philadelphia police that sat unnoticed for two months, and companies testified in New York that they can’t say how likely their systems are to cause harm. Right now neither party has earned much trust on this, and the 39% who trust no one are telling both sides the same thing: show us real rules.
The open question: will either party run on AI safety rules before the midterms, or keep leaving it to the companies?
Also on BeezLoop: an Anthropic AI model sent Philadelphia police a fake murder tip, and AI companies testified at NYC City Hall.
Sources: AP-NORC Center for Public Affairs Research · Scripps News (video)
