Americans Are Growing More Worried About AI. Nearly 3 In 4 Say Tech Firms Aren’t Doing Enough To Prevent Harm.

Americans are becoming more concerned about artificial intelligence as companies roll out increasingly capable systems, with nearly three-quarters saying AI firms have not done enough to prevent the technology from causing serious harm to society, according to a new survey.

About 73% of U.S. adults said they were worried that AI companies had not gone far enough to prevent serious harm, a Reuters/Ipsos poll released showed.

Negative views of the technology have also increased. About 39% of respondents said AI was having a negative impact on society, up from 36% last month and the highest level since the poll began asking the question in March. Only 11% said AI was having a positive impact, while the remaining respondents were unsure or did not answer.

The poll also found support for a slower pace of development. About 55% of respondents said slowing AI development would be a good thing, compared with 13% who said it would be bad. The remaining respondents were unsure or did not answer.

Americans also placed greater importance on safety than maintaining the country's lead in AI. About 73% said ensuring that AI is developed safely and responsibly was more important, compared with 23% who prioritized keeping the U.S. ahead of other countries in AI development.

The survey included 1,277 adults across the U.S. and had a margin of error of about 3 percentage points.

Several major AI companies and researchers have publicly discussed the risks posed by increasingly capable systems. About 69% of respondents said they had followed reports about calls from AI companies for slower development, greater oversight and stronger safety standards.

The incidents have added to a broader discussion inside the AI industry over safety controls. Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and other industry figures have recently supported measures aimed at slowing some forms of advanced AI development and establishing stronger safeguards.

The debate intensified after an Anthropic researcher resigned this month over concerns about advanced AI. Anthropic scientist Evan Hubinger separately said he believed there was a greater than 10% chance that AI could cause human extinction within the next decade. Researchers and technology executives disagree widely over the likelihood and timing of such extreme outcomes.

The poll also found limited confidence in leaving AI oversight entirely to technology companies, with most respondents saying federal officials should play a major role in establishing safety standards.

Calls for federal action have come from figures across the political spectrum. Sen. Bernie Sanders, an independent who caucuses with Democrats, and Steve Bannon, a longtime ally of President Donald Trump, jointly called this month for restrictions on advanced AI development, including federal safety standards.

AI safety has also become part of discussions between Washington and Beijing. Treasury Secretary Scott Bessent said Monday that the U.S. proposed an incident-notification system during talks with Chinese officials that would allow the two countries to alert each other about AI events involving national security risks.

The talks came ahead of meetings between Trump and Chinese President Xi Jinping in Washington this week. U.S. and Chinese officials have continued discussing AI alongside trade and other economic issues as companies in both countries develop increasingly powerful models.

Meanwhile, the U.N. Security Council is scheduled to discuss artificial intelligence Wednesday, with representatives from OpenAI, Anthropic and Chinese AI companies expected to participate. Chinese startup DeepSeek has been invited to brief the council as governments discuss the security implications of advanced AI.

Concerns about AI have grown alongside its expanding use in workplaces. A separate Reuters/Ipsos survey in June found that 53% of Americans worried that AI could cause them or someone in their household to lose a job, while 37% said they were not worried about that possibility.