Most Americans Say AI Is Bad for Society, Oppose More Data Centers

A national poll shows a majority of Americans view artificial intelligence as harmful to society: 57% say AI is bad versus 42% who say it is good, with pessimism strongest among younger adults (65% of those under 30). Partisan divides are clear - 63% of Democrats and 58% of independents call AI bad for society, while a slim majority of Republicans see it as good. People are more mixed about AI’s personal impact (50% good, 49% bad), but pessimistic intensity is higher (15% say AI is extremely bad for them vs. 3% extremely good). Respondents judge many AI outcomes as high risk: 73% cite hacking, 65% fear job losses, 62% worry people will lose decision-making autonomy, and 54% expect greater economic inequality; 33% even view human extinction as a high risk. Fewer see big benefits - 34% expect easier everyday tasks, 30% expect U.S. dominance in AI, 20% expect greater productivity, and only 13% expect many new jobs.

Two-thirds of Americans say the federal government is not doing enough to regulate AI, and 66% oppose building more data centers in the U.S. (72% oppose one in their community). Opposition spans parties and demographics - 81% of Democrats, 76% of independents and 58% of Republicans oppose local data centers - with higher resistance among young adults and rural residents. Approval of how the president is handling AI is low overall (30% approve), though approval is high among MAGA-aligned Republicans. The poll surveyed 2,545 U.S. adults Oct. 9-13, 2026, with a roughly ±2-point margin of error.