Most Americans Say AI Companies Are Not Doing Enough to Prevent Major Risks
WASHINGTON — September 23, 2026

A large majority of Americans say artificial intelligence companies are not doing enough to prevent potentially dangerous consequences from increasingly powerful AI systems, according to a new Reuters/Ipsos poll.

The survey found that 73% of respondents were concerned that AI companies were not taking sufficient steps to prevent potentially catastrophic outcomes from the technology.

Concern About AI Is Rising
The poll found that 39% of Americans now believe AI is having a negative impact on society, the highest level recorded since the Reuters/Ipsos tracking began in March.

Only 11% said they viewed AI’s impact on society positively.

The results suggest that while AI continues to attract investment and attention for its potential benefits, concerns about its risks are becoming increasingly widespread among the public.

Many Want Stronger Oversight
The survey also found limited confidence in the ability of AI companies to regulate themselves.

A majority of respondents supported federal government involvement in establishing safety standards for AI development.

At the same time, the poll found that Americans remain interested in the potential benefits of the technology, including its possible contributions to science and the economy.

Debate Over How Quickly AI Should Develop
One of the most significant findings was that 55% of respondents supported slowing down AI development in some way.

The poll also found that 73% prioritized safe and responsible AI development over maintaining global competitiveness in the technology sector, while 23% placed greater emphasis on competitiveness.

The findings come as technology companies continue to develop increasingly advanced AI systems and governments debate how they should be regulated.

Cybersecurity and Other Risks

Concerns highlighted in the survey include cybersecurity threats and the possibility that increasingly autonomous AI systems could cause serious harm.

The Reuters/Ipsos poll was conducted online over four days among 1,277 U.S. adults and has a margin of error of about three percentage points.

The results provide a snapshot of public attitudes in the United States at a time when AI development is accelerating and companies are competing to build more capable systems.

As the technology advances, the survey indicates that many Americans are increasingly focused not only on what AI can accomplish, but also on how its risks should be controlled.

Source: Reuters/Ipsos.