A Pew Research Center survey of 5,023 U.S. adults found that 57 percent rate the risks of artificial intelligence for society as high or very high, while 25 percent rate the benefits as high or very high. Statista charted the results in an infographic published February 27, 2026. Pew fielded the survey from June 9 to 15, 2025.

Among respondents who see high risks, the most cited reason was that AI erodes human abilities and connections, named by 27 percent. Concern about the accuracy of information followed at 18 percent, and concern over human control of AI came next at 17 percent. AI misuse for bad or criminal purposes drew 11 percent, and job losses ranked fifth at 9 percent.

Respondents who see high benefits pointed most often to efficiency. Of that group, 41 percent said AI makes people more efficient, 23 percent said it expands human and technological abilities, and 7 percent cited general societal benefits.

The Statista write-up described a gap between the public view and the stance of AI industry leaders. It noted that Nvidia chief executive Jensen Huang and OpenAI chief executive Sam Altman had recently expressed surprise at the level of skepticism and at how slowly enterprises were adopting the technology.

The chart excludes respondents who answered not sure. Because the sample covers U.S. adults, the percentages describe American opinion and give a baseline for how the public weighs AI risk at a time when companies are reporting documented incidents involving data exposure and system errors.

Source: Statista - https://www.statista.com/chart/35906/perceived-risks-and-benefits-of-ai-for-society/