Two and a half years after ChatGPT turned "generative AI" into a dinner-table phrase, the technology has quietly crossed into the American mainstream. About half of US adults now report using an AI chatbot, according to a sweeping new Pew Research Center survey released June 17, 2026 — and roughly one in four say they reach for one every single day. Yet the same Americans who are folding chatbots into their searches, their workdays and even their doctor's-office questions remain deeply unconvinced that the technology is good for the country. Adoption is up. Trust is not.
That gap — between what people do with AI and what they believe about it — is the central tension running through Pew's "Americans and AI 2026" report.
What the survey found
Pew surveyed 5,119 US adults from Feb. 17–23, 2026, through its nationally representative American Trends Panel. The headline number depends on how you slice it. About 49% of US adults say they use AI chatbots, up from roughly a third of respondents in Pew's 2024 reading — a jump of more than 15 percentage points in two years. A separate, narrower figure appears in much of the coverage: 44% of adults specifically use ChatGPT, which remains the single most-used chatbot. The two numbers are easy to conflate but measure different things; the 49% captures chatbot use broadly, while 44% is OpenAI's individual product reach.
ChatGPT's dominance is striking. Google's Gemini ranked second at 24% of adults, followed by Microsoft Copilot (17%), Meta AI (14%), Elon Musk's Grok (8%), Anthropic's Claude (6%) and Character.AI (3%). Notably, ChatGPT is now used by similar shares of men and women — a reversal from two years ago, when men led women in overall chatbot use by 11 points.
What are people actually doing with these tools? The most common use, cited by 42% of users, is searching for information. After that: using chatbots for fun (25%), creating or editing images and video (24%), seeking medical advice (20%) and getting diet or fitness information (20%). Smaller shares turn to chatbots for news (13%), emotional support (10%) or companionship (4%). AI features are also seeping into hardware: 37% of adults report using AI-equipped smartwatches and 35% use smart speakers such as an Amazon Echo or Apple HomePod.
The trust problem
For all that uptake, Americans' mood toward AI tilts notably negative. Roughly two-thirds believe AI is advancing too fast. Forty percent think it will ultimately be worse for society. And confidence in the institutions meant to manage the technology is thin: 67% of Americans have little or no confidence that the US government can effectively regulate AI, while 59% say the same about whether US companies will develop it responsibly. On privacy, the verdict is nearly uniform — 71% agree that AI will make their personal information less secure.
"AI is no longer the future; for many, it's here and now," Jeffrey Gottfried, an associate director of research at Pew Research Center, said in a statement accompanying the report. "Americans are increasingly using chatbots and bringing AI into their homes, but they have a complex relationship with AI. They may use it, but they're still highly skeptical of it and how it will impact our society."
That skepticism is not limited to older or non-technical Americans. Pew found that even younger adults — by far the heaviest users — lean negative on how quickly AI is advancing. Adults under 50 are far more likely to use chatbots than those 50 and older, but enthusiasm for the tools does not translate into trust in their trajectory.
The divides beneath the average
The aggregate "half of Americans" figure hides sharp demographic splits. The clearest is age: under-50 adults dominate chatbot use, while older Americans remain comparatively absent. Among the 51% of US adults who don't use AI chatbots, the most common reason — cited by about 60% — is simply a lack of interest, not fear or lack of access. That detail matters for anyone betting on universal adoption: a large bloc of non-users isn't waiting to be convinced about safety so much as shrugging.
Then there's gender. The headline gap in whether people use chatbots has essentially closed, but the gap in how they use them has not. Men report using chatbots more regularly — 27% daily versus 20% of women — and are more likely to deploy them across a range of tasks, including work. Pew also found women are consistently more skeptical about AI, including its effect on their own lives. In other words, parity in access masks a persistent divide in intensity and attitude.
Analysis: living with a tool you distrust
The defining feature of the 2026 AI landscape, on Pew's evidence, is cognitive dissonance at scale. Americans are not rejecting AI — they are absorbing it into daily routines while withholding their endorsement. They use ChatGPT to draft an email at 9 a.m. and tell pollsters at 9 p.m. that the technology is moving too fast and that no one in Washington can be trusted to rein it in.
This has real consequences for the policy debate. Public appetite for regulation is high, but so is public pessimism that regulation will work — a combination that gives lawmakers a mandate without confidence. It is also a warning to AI companies: market penetration is not the same as goodwill. Reaching half the country has not bought the industry trust, and the privacy numbers (71% expecting less security) suggest the reputational deficit is concentrated precisely where these firms make their money — personal data.
The medical-advice finding deserves its own flag. One in five chatbot users now consults AI for health guidance, a behavior that outpaces both the tools' reliability guarantees and any regulatory framework. As adoption normalizes, the stakes of hallucinated or biased outputs rise from inconvenient to potentially dangerous.
What to watch
Three things will signal whether this uneasy equilibrium holds. First, the daily-use number — currently around one in four — is the truest measure of habit; if it climbs toward the headline adoption figure, AI will have moved from novelty to infrastructure. Second, the regulation gap: with two-thirds of Americans doubting Washington's competence even as they demand oversight, watch whether any 2026 legislative push can close the credibility deficit Pew documented. Third, the gender and age intensity gaps — if heavy use stays concentrated among younger men, the "everyone uses AI now" narrative will remain, at best, half true.
For now, the American verdict on AI is captured in a single contradiction: indispensable, and not quite trusted.
"Americans are increasingly using chatbots and bringing AI into their homes, but they have a complex relationship with AI. They may use it, but they're still highly skeptical of it and how it will impact our society."-- Jeffrey Gottfried, Associate Director of Research, Pew Research Center