Most Americans Are Skeptical of AI, Pew Survey Finds
Only 16 percent of Americans believe artificial intelligence has a positive effect on society, according to a Pew Research Center survey released this week, a finding that lands at an awkward moment for an industry in the middle of its biggest public-market run. The survey found that a majority of Americans hold neutral or negative views of AI’s long-term social impact, even as their use of the technology continues to rise. The gap between adoption and trust is now one of the defining facts of the AI economy, and companies are beginning to feel its consequences.
The Pew findings cut across the industry’s own narrative. Technology executives have spent two years describing AI as a productivity revolution that will reshape work, medicine and education, and the summer’s IPO wave has given that story a financial vote of confidence. The public is not buying it. Respondents expressed skepticism about AI’s effect on jobs, on the quality of information, and on society at large, with concerns about automation and misinformation outweighing hopes about convenience and capability.
The disconnect between usage and sentiment is the survey’s most striking result. Americans are using AI tools in rising numbers, for writing, coding, research and routine tasks, and the share of adults who say they have tried the technology has climbed steadily. But familiarity has not bred comfort. The people who use AI most are often the ones most aware of its limitations, researchers said, and everyday exposure to flawed answers, fabricated citations and intrusive data collection may be reinforcing skepticism rather than dispelling it.
For the largest AI companies, the numbers represent both a brand problem and a policy problem. A public that distrusts a technology is a public that will tolerate regulation of it, and the industry’s opponents have seized on the gap to argue for tighter oversight. Lawmakers on both sides have proposed rules governing AI’s use in hiring, health care and policing, and the survey gives them evidence that their constituents share the concern. Companies that move too aggressively to commercialize AI, the argument goes, invite the regulatory backlash that follows unpopular industries.
The industry’s response has been to argue that the technology is new and that benefits will accumulate faster than harms. Executives point to falling prices for AI services, expanding capabilities and early evidence of productivity gains in fields from software to law. They also note that public opinion about new technologies typically improves with time, citing the trajectory of the internet and smartphones, which were greeted with skepticism before becoming indispensable. Whether AI follows that path depends on whether its visible failures become rarer, a process that is far from guaranteed.
Some companies are already adjusting their behavior in response to the mood. Product teams have added disclaimers, guardrails and human-review options, partly to reduce the errors that feed public distrust. Marketing departments have toned down the most grandiose claims, and executives have begun talking about AI as a tool to augment workers rather than replace them. The shifts are incremental, but they reflect a recognition that the technology’s biggest obstacle is no longer its capability, which improves monthly, but its reputation, which has barely moved.
The survey also found meaningful variation beneath the headline number. Younger Americans are more positive about AI than older ones, and people who work with the technology professionally are more favorable than those who encounter it only as consumers. Partisanship plays a role as well, with the details of how different groups answered reflecting the broader political divisions of the country. But in no group did a majority express the enthusiasm that the industry’s marketing implies, and in several the negative views were dominant.
For the industry, the implication is that the race to build AI is also a race to be trusted, and the two are not moving at the same speed. Capital markets have priced AI companies for a future in which the technology is everywhere; the public’s view suggests that future will require persuasion, regulation and time. The companies that manage the transition best, analysts said, will be those that treat public opinion as a constraint as real as compute costs or chip supply. The Pew numbers show how far that work has to go. The survey’s methodology also points to the stability of the sentiment. Pew has asked similar questions about AI over the past several years, and while the wording has varied, the overall shape of the results has not changed much: a minority positive, a minority negative, and a large middle ground that leans skeptical. That stability is notable because the technology itself has improved dramatically over the same period, with models getting faster, cheaper and more reliable. If capability alone could move public opinion, the numbers should have moved more. Their refusal to budge suggests the distrust is rooted less in what the technology does today than in how it is perceived to treat the people who encounter it.
This article was prepared by Rhino Finance’s editorial team based on public reporting.
