Multiple Polls Find Most Americans Think AI is Moving Too Fast

Tech Policy Press

Source published: 13 Oct 2026, 18:11 · TehranSite published: 13 Oct 2026, 18:37 · Tehran

Protesters demonstrate outside of the OpenAI DevDay 2026 conference, Tuesday, Sept. 29, 2026, in San Francisco. (AP Photo/Jeff Chiu)

Following multiple incidents over the past several months where AI agents operated by OpenAI, Anthropic, and Meta intruded on third party systems, industry leaders including Anthropic’s Dario Amodei called for moderating the pace of AI development. But the Trump administration has rejected calls for a slowdown.

“We’re leading China in AI, we’re the most sophisticated country in the world, and frankly, I want to keep it that way because whoever wins AI, wins,” President Donald Trump told reporters last month. “And we can put guardrails, we can do this and that. But I think you have a lot of negative forces that are bringing it up that shouldn’t be bringing it up.”

In five big national polls in 2026, the share of American adults saying AI is moving too fast or should slow down ranged from 55% to 77%. Each poll worded the question differently, as shown below. In the polls that broke out results by party, Democrats were more concerned about AI than Republicans. But in three of the four polls that asked by party, a majority of Republicans also said AI was moving too fast.

The Washington Post reported that Democrats have “seized on” the Trump administration’s resistance to AI regulation “as a line of attack” in the US midterm elections.

All respondentsBy party

Share saying AI is moving too fast, or that companies should slow down or stop, by party, where the pollster published it.

The Economist/YouGov
71% advancing too quickly
71%21%
“Do you think artificial intelligence (AI) is advancing too quickly or too slowly?”
Oct. 2-5, 2026 · 1,381 registered voters · margin of error ±3.4 percentage points · Source
AP-NORC
64% developing too fast
64%27%8%
“Do you think artificial intelligence (AI) technology in the U.S. is developing too fast, not developing fast enough, or developing at about the right place?”
“Right place” appears as printed in AP-NORC’s topline.
Sept. 24-28, 2026 · 2,140 U.S. adults · margin of error ±2.9 percentage points · Source
Quinnipiac University
77% say stop or slow down
30%47%14%
“Which comes closest to your line of thinking, AI companies in the U.S.: should stop developing powerful AI systems until their safety can be evaluated, should slow down developing powerful AI systems until their safety can be evaluated, should keep developing powerful AI systems at the current pace, or should move faster in developing powerful AI systems?”
Sept. 24-27, 2026 · 1,202 U.S. adults · margin of error ±3.5 percentage points · Source
CBS News/YouGov
55% say stop or slow down
12%43%37%
“What should U.S. A.I. companies do with regard to the speed of A.I. research and development? Do you think they should...”
Sept. 16-18, 2026 · 2,051 U.S. adults · margin of error ±2.5 percentage points · Source
Fox News
59% moving too quickly
59%33%
“Do you think the use of artificial intelligence is moving too quickly, too slowly, or at about the right pace in the United States?”
Jan. 23-26, 2026 · 1,005 registered voters · margin of error ±3 percentage points · Source

The Economist/YouGov
71% advancing too quickly
Democrats
82%
Republicans
59%
Independents
74%
“Do you think artificial intelligence (AI) is advancing too quickly or too slowly?”
Oct. 2-5, 2026 · 1,381 registered voters · margin of error ±3.4 percentage points · Source · Party figures
AP-NORC
64% developing too fast
AP-NORC’s topline counts independents who lean toward a party with that party; it does not report a figure for other independents on this question.
Democrats
76%
Republicans
54%
Independents
Leaners counted with parties
“Do you think artificial intelligence (AI) technology in the U.S. is developing too fast, not developing fast enough, or developing at about the right place?”
“Right place” appears as printed in AP-NORC’s topline.
Sept. 24-28, 2026 · 2,140 U.S. adults · margin of error ±2.9 percentage points · Source · Party figures
CBS News/YouGov
55% say stop or slow down
Democrats
68%
Republicans
41%
Independents
58%
Darker segment is “stop development”; totals add the two rounded figures.
“What should U.S. A.I. companies do with regard to the speed of A.I. research and development? Do you think they should...”
Sept. 16-18, 2026 · 2,051 U.S. adults · margin of error ±2.5 percentage points · Source
Fox News
59% moving too quickly
Democrats
64%
Republicans
54%
Independents
60%
“Do you think the use of artificial intelligence is moving too quickly, too slowly, or at about the right pace in the United States?”
Jan. 23-26, 2026 · 1,005 registered voters · margin of error ±3 percentage points · Source · Party figures
Quinnipiac University
77% say stop or slow down
Quinnipiac did not publish a party breakdown for this question.
“Which comes closest to your line of thinking, AI companies in the U.S.: should stop developing powerful AI systems until their safety can be evaluated, should slow down developing powerful AI systems until their safety can be evaluated, should keep developing powerful AI systems at the current pace, or should move faster in developing powerful AI systems?”
Sept. 24-27, 2026 · 1,202 U.S. adults · margin of error ±3.5 percentage points · Source

Wary of AI’s Effects

Each pollster asked its own question again, a year or more apart, so change can be compared.

Share who are more concerned than excited, or who say AI will do more harm, by party. Quinnipiac’s party figures come from its March 2026 poll.

Pew Research Center
52% more concerned than excited
52%37%9%
“Overall, would you say the increased use of artificial intelligence (AI) in daily life makes you feel…”
June 22-28, 2026 · 3,488 U.S. adults · margin of error ±1.8 percentage points · Source
Same question, earlier: About the same as in June 2025, when it was 50%.
June 9-15, 2025 · U.S. adults · sample size and margin of error not listed in the topline
Quinnipiac University
53% say AI will do more harm
53%34%13%
“When it comes to your day-to-day life, do you think AI will do more good than harm or do more harm than good?”
Sept. 24-27, 2026 · 1,202 U.S. adults · margin of error ±3.5 percentage points · Source
Same question, earlier: Up 9 percentage points since April 2025, when it was 44%.
April 3-7, 2025 · 1,562 U.S. adults · margin of error ±2.5 percentage points · Source
Quinnipiac also asked this question in March 2026, when 55% said “do more harm.” The rise came between April 2025 and March 2026; Quinnipiac says the September result is little changed from March.

Pew Research Center
52% more concerned than excited
Pew counts independents who lean toward a party with that party and does not give a figure for other independents.
Democrats
56%
Republicans
49%
Independents
Leaners counted with parties
“Overall, would you say the increased use of artificial intelligence (AI) in daily life makes you feel…”
June 22-28, 2026 · 3,488 U.S. adults · margin of error ±1.8 percentage points · Source · Party figures
Same question, earlier: About the same as in June 2025, when it was 50%.
June 9-15, 2025 · U.S. adults · sample size and margin of error not listed in the topline
Quinnipiac University
53% say AI will do more harm
The party figures below come from Quinnipiac’s March 19-23, 2026, poll of 1,397 adults, when 55% overall said AI would do more harm. Quinnipiac did not break down its September result by party.
Democrats
59%
Republicans
52%
Independents
58%
“When it comes to your day-to-day life, do you think AI will do more good than harm or do more harm than good?”
Sept. 24-27, 2026 · 1,202 U.S. adults · margin of error ±3.5 percentage points · Source · Party figures
Same question, earlier: Up 9 percentage points since April 2025, when it was 44%.
April 3-7, 2025 · 1,562 U.S. adults · margin of error ±2.5 percentage points · Source
Quinnipiac also asked this question in March 2026, when 55% said “do more harm.” The rise came between April 2025 and March 2026; Quinnipiac says the September result is little changed from March.

About these polls

Each row is one poll question. The questions are worded differently, so compare the polls by their wording, not as a trend. Change notes compare the same pollster’s same question at two points in time.