A new poll finds Americans are increasingly incorporating artificial intelligence into their daily lives while remaining deeply concerned about where the technology is headed, with broad opposition to AI data centers in local communities and widespread doubts about the technology's effects on jobs, misinformation and society.

The NBC News Decision Desk survey, conducted online among 7,105 adults from Aug. 20 through Sept. 1, found that 70 percent of Americans are more worried about artificial intelligence than excited about it. At the same time, 52 percent said they use AI tools either very often or sometimes, a six-point increase from a similar survey conducted in June 2025. The share who said they use AI “very often” rose to 23 percent, compared with 14 percent a year earlier.

The growing use of AI has not translated into broad public support for the infrastructure needed to power the technology. Sixty-nine percent of respondents said they oppose the construction of an AI data center in their local area, including 45 percent who strongly oppose such facilities. Only 31 percent said they support building an AI data center nearby.

The opposition comes as data centers have become an increasingly visible issue in communities across the country. The facilities can generate jobs and tax revenue, but residents have raised concerns about their demands for electricity and water, as well as noise, land use, and other potential effects on surrounding communities. Opposition has emerged across traditional political lines, making AI data centers an increasingly prominent issue heading into the 2026 midterm elections.

The poll also found that 81 percent of Americans believe the government is not doing enough to regulate artificial intelligence. About two-thirds said AI poses a high risk to society, while only 18 percent said they can trust information generated by AI all or almost all of the time.

Neither political party has established a clear advantage with voters on AI policy. Forty-four percent said they trust neither Democrats nor Republicans to handle the issue. Twenty percent said they trust Democrats, while 16 percent chose Republicans. Another 15 percent were unsure, and 5 percent said they trust both parties equally.

Americans also remain divided over who should be responsible for protecting consumers from AI-generated misinformation. Thirty-eight percent said the government should take the lead, while 34 percent placed the responsibility on the companies developing AI tools. Another 28 percent said individuals themselves should be primarily responsible.

Concerns about AI extend across several major areas of American life. Majorities believe the technology will do more harm than good in K-12 education, news and information, elections and democracy, higher education and national security. Pluralities also expect AI to cause more harm than good in the economy, health care, and arts and entertainment.

Misinformation ranks among the public's biggest concerns. Nearly three-quarters of respondents cited deepfakes, fraud, and misinformation as major worries, while roughly two-thirds said they are concerned that Americans are becoming too dependent on artificial intelligence.

Jobs are another significant source of concern. Seventy percent of Americans said AI is taking away job opportunities, compared with just 7 percent who said it is creating opportunities. Twenty-three percent said they either saw no difference or were unsure.

Younger Americans expressed particularly strong concerns about the employment impact. Among adults ages 18 to 29, 73 percent said artificial intelligence is costing people job opportunities. The concern also varied by political affiliation, with 78 percent of Democrats and 69 percent of independents expressing concern about AI reducing employment opportunities, compared with 62 percent of Republicans.

Despite the broader unease, Americans see potential benefits from AI in certain areas. Fifty-five percent said they believe the technology will do more good than harm in scientific research, while 26 percent said the opposite. Forty-seven percent also expect AI to provide more benefits than harm in health care, compared with 30 percent who expect more harm.

The survey found a gender divide on some of those questions. Nearly two-thirds of men said they expect AI to benefit scientific research, compared with 45 percent of women. Men were also twice as likely as women to say AI will help the economy more than it hurts it.

Americans' willingness to rely on AI also depends heavily on the task. Nearly three-quarters said they would feel comfortable relying on AI-generated recipes or cooking instructions. That confidence drops substantially when the subject involves historical events or health conditions and medical treatments, with only about one-third expressing confidence in AI-generated information on those subjects.

Only 21 percent said they would be comfortable relying on AI-generated information for breaking news or personal financial decisions.

The polling comes as the rapid expansion of artificial intelligence and the data centers supporting it becomes a growing national and local political issue. While technology companies and supporters argue that data centers can bring jobs, investment, and tax revenue, opponents across the political spectrum have raised concerns about their impact on electricity demand, water supplies, land, and local communities.