AI has an optics problem. For years, surveys have shown that public perception of AI is largely negative. When the Pew Research Center polled thousands of U.S. adults in 2024, only 17% of them thought the technology would have a positive impact in the years to come. Most of them were especially concerned with job losses, with 64% of respondents saying they believed AI would eliminate jobs in the near future.
A Gallup poll this week indicates that Americans have actually grown more skeptical of AI, even as they have become more familiar with the technology. While over half of Americans (52%) still think AI can do equal amounts of good and harm, there’s been a sharp increase in the number of people who have fallen out of love with AI.
In 2024 and 2025, Gallup found that 31% of people surveyed said they felt AI did more harm than good. This year, that figure jumped to 39%—and among young people between the ages of 18 and 29, it was even higher at 47%. Only 9% of people thought AI could do more good than harm, down from 12% last year.
AI adoption has climbed steadily in the workplace and beyond, suggesting the shift in public opinion is informed by a greater awareness of AI’s potential for harm, rather than fear of the unknown. Drawing on data from a nationally representative household survey, the Federal Reserve found earlier this year that 41% of employees now use generative AI in the workplace, while half of Americans use it outside of work (an increase of 31% and 26% year over year, respectively). In the Gallup survey, 70% of respondents said they are somewhat or extremely knowledgeable about AI, as compared with 64% two years prior.
Despite being more comfortable with the technology, workers are still worried about how it could impact their livelihood—in part because of how the business world has enthusiastically embraced it. According to Gallup, many Americans have little faith that companies will use AI responsibly, with only 27% of respondents saying they have at least some trust in businesses, a decrease from 31% last year.
Much of their skepticism also seems driven by the belief that AI will kill jobs, even if that hasn’t been borne out yet in overall employment data. (Some employers even claim AI will spur them to hire more people.) Nearly eight in 10 people polled by Gallup said AI would slash jobs over the next decade, and there was a significant uptick among older and younger respondents relative to their concerns about job loss two years ago.
Many of these fears may have been stoked in the last year as major companies have repeatedly cited AI in their layoff announcements, even while some of them have insisted it’s not the primary reason. AI is likely just one of a constellation of factors behind corporate restructuring decisions—but that may be little comfort for workers who have more existential concerns about how AI could fundamentally alter their lives.