Majority Of Americans Say AI Does More Harm Than Good, Poll Finds
That shift comes despite the fact that most respondents are already using AI in some capacity. Only 27% say they've never used it at all, while more than half report using AI for research, a sharp increase from 37% a year ago. Other use cases, including data analysis and image generation, are also on the rise. Notably, about one in five respondents say they've turned to AI for medical advice, which is an eyebrow-raising figure given the ongoing concerns around AI reliability.
Indeed, reliability is exactly where skepticism remains firmly entrenched. A full 76% of those surveyed say they are skeptical of AI's accuracy, a number unchanged from last year. The disconnect between widespread usage paired with persistent distrust runs through much of the data. For example, 64% of respondents say AI is harmful for education, even as 51% report using it for research purposes.
More broadly, respondents express significant anxiety about the technology's societal impact. Roughly 62% say they are not excited about AI's future, while 80% report being at least somewhat concerned about its effects. Job displacement remains a key fear, with 70% believing AI will lead to fewer jobs overall. Interestingly, however, 69% of employed respondents say they are not especially worried about losing their own jobs to AI.
There are also signs of a trust gap when it comes to institutions. Some 76% of respondents believe companies are not transparent enough about their use of AI, and 74% say the government isn't doing enough to regulate it, with both figures ticking upward from last year. At the same time, 65% oppose the construction of local AI data centers, suggesting resistance not just to the technology itself, but to the infrastructure supporting it.
Healthcare offers another telling data point, as 81% of respondents say they would still want a human doctor involved in their care even in the case that an AI system was proven more reliable. Meanwhile, 56% claim they are confident they can distinguish between real and AI-generated video. That figure is downright comical given the current state of generative media online; mega-popular accounts endlessly post shamelessly unmarked AI slop that gets tens of millions of views and hundreds of thousands of reposts. If people can tell the difference, they certainly don't care.
The poll also highlights generational divides, although they're not as broad as you might think in most cases. Younger respondents, particularly Gen Z, generally report higher levels of familiarity with AI, but also greater pessimism about its long-term impact, especially on employment. One of the starkest splits appears around military use, where overall opposition sits at 69%, but nearly half of older respondents express support of military use of AI. This tracks with broad generational opinions on war and the military in general, though.
Chetan Jaiswal, Ph.D., Associate Chair of Computing at Quinnipiac University, summarized the findings succinctly: "Americans are not rejecting AI outright, but they are sending a warning. Too much uncertainty, too little trust, too little regulation, and too much fear about jobs." Of course, as with many public opinion polls, it's worth keeping the sample size in mind. With just under 1,400 respondents, the findings offer a snapshot of broader sentiment, not a definitive summary. Still, the trends are hard to ignore: AI adoption is accelerating, but confidence in the technology is not keeping pace.
Top photo by Tima Miroshnichenko.
