Almost half of American adults now use AI chatbots, but only 16 percent think the technology will have a positive impact on society—and most expect AI to make society worse. That's according to a Pew Research Center survey which found chatbot use soaring, with roughly a quarter of U.S. adults using the tools daily. About a third of adults knowingly bought AI-enhanced gadgets, notwithstanding the shaky truthfulness of marketing claims around the technology.
But the respondents were deeply skeptical about its impact, and the younger they were the more likely they were to have negative opinions.
Pew surveyed 5,119 U.S. adults in February through its American Trends Panel, a randomly recruited group built to mirror the adult population. It seems unlikely that these trends have changed direction in the last four months; the gap between use and approval was the survey's key finding.
40 percent of those polls said the impact will be generally negative. On personal effects, 31% expected harm and 23% expected benefits. About two-thirds said AI is advancing too quickly, and roughly seven in 10 said it will make their personal information less secure. Adults under 30 were the most negative on society, at 48%, despite using the tools most.
The most popular chatbots, such as OpenAI's ChatGPT, Anthopic's Claude and Google's Gemini, are the money-burning killer app as far as everyday Americans are concerned. They run on large language models, trained on enormous volumes of text so as to generate more text in response to prompts and capable of drafting, summarizing and jabbering convincingly. ChatGPT is the dominant platform, used by 44% of those paneled by Pew.
The "frontier models" pushing those abilities, the largest and most capable systems, recently began attracting government scrutiny. On June 12, the Commerce Department ordered Anthropic to cut off foreign access to its Fable 5 and Mythos 5 models, citing national security. Unable to filter users by nationality, the company disabled both worldwide. OpenAI yesterday said ChatGPT 5.6 was ready, but that only invited users would get access to it for now.
If dismay at AI comes in various forms, corporate America's eagerness to blame it for mass layoffs is chief among them. Outplacement firm Challenger, Gray & Christmas attributed 21,490 April layoffs to AI and automation. But analysts cited by the Society for Human Resource Management question AI's role in the workforce reduction: the layoffs are subject to "AI-washing" to reframe cost-cutting driven by pandemic overhiring and C-suite ineptitude.
Remember the face up top? Microsoft's Holocaust-denying Tay chatbot was more than a decade ago.
