A Reuters/Ipsos poll of 1,277 U.S. adults found 73% say AI companies have not done enough to prevent serious societal harm. Only 11% see AI’s impact on society as positive, while 39% view it negatively, the highest share in the poll’s tracking since March 2026.
The poll also found 55% say slowing AI development would be good and 13% say it would be bad. Meanwhile, 73% prioritize safe, responsible development over keeping the U.S. ahead of other countries; 23% prioritize competitiveness. Most respondents said federal officials should play a major role in setting AI safety standards.
Conducted online and completed Sept. 20, the poll was published Sept. 22 and has a margin of error of plus or minus three percentage points. It measures respondents’ perceptions, not the effectiveness of companies’ safety programs; respondents did not specify a preferred agency or enforcement mechanism.
