Only 16% of Americans believe in the positive impact of AI on society — a large-scale survey reveals a troubling trend.
American society is demonstrating an unprecedented level of skepticism toward artificial intelligence. According to a large-scale survey, only 16% of U.S. adults expect AI to have a positive impact on society. Meanwhile, nearly 40% of respondents are confident the effect will be negative.
The study, which involved 5,119 participants, revealed deep distrust not only of the technology itself but also of the institutions meant to regulate it. 67% of respondents do not believe the U.S. government can effectively regulate the AI sector. Another 59% distrust the developing companies. Nearly two-thirds of respondents believe that the development of artificial intelligence is happening too quickly.
The breakdown by political preference is particularly telling. The greatest skepticism is shown by young Democratic voters under 30 — among them, only 14% expect a positive impact from AI on society. This suggests that even traditionally progressive segments of the population are beginning to view the technology as a threat.
Notably, against the backdrop of general pessimism, there is a sharp increase in the everyday use of neural networks. About a quarter of Americans reported using chatbots daily. ChatGPT remains the undisputed leader — 44% of U.S. adults use it, more than double the figures from 2023. It is followed by Gemini (24%), Copilot (17%), and Meta AI (14%).
These findings are corroborated by other studies. In May, a YouGov poll showed that 71% of Americans consider AI development too fast, with pessimists outnumbering optimists by a ratio of 2:1 (51% vs. 25%). For comparison, as recently as January 2025, the picture was more balanced — 35% expected a positive impact, while 34% expected a negative one.
Recall that in a joint Reuters and Ipsos study in May 2025, 61% of Americans called AI a direct threat to humanity.
My analyst comment: We are witnessing a classic paradox of breakthrough technology adoption: mass acceptance in daily life combined with deep distrust at the systemic level. Users actively employ AI tools for everyday tasks, yet fear their long-term consequences. This creates a unique market situation where demand for AI products is growing, but any tightening of regulation or a high-profile incident could trigger a sharp decline in trust. Investors and developers should take this fragile balance into account.