Where AI Public Sentiment Stands After Pew's 2026 Poll

The gap between the stock market and the street has rarely looked this wide. While investors keep pouring money into AI, AI public sentiment among ordinary Americans is sitting at a notably grim low. A Pew Research Center survey released on June 17, 2026, found that only 16 percent of Americans expect AI to have a positive impact on society over the next two decades, against 40 percent who expect a negative one. A separate consumer study published the same week added a commercial twist, reporting that most shoppers now treat the word AI in marketing as a reason to look away rather than lean in.

What the Pew numbers actually say

The survey is large and recent, which matters when a single statistic gets quoted everywhere. Pew polled 5,119 US adults from February 17 to 23, 2026, through its American Trends Panel, a sample built to represent the full adult population. As TechCrunch summarized it, the contrast with Wall Street enthusiasm is the whole story. The people building and funding AI see a windfall, and the people living alongside it mostly see something to worry about.

The pessimism runs deeper than the headline figure. Pew found that 63 percent of Americans say AI is advancing too quickly, and a mere 2 percent think it is moving too slowly. On personal stakes, 71 percent believe AI will make their personal information less secure. When the question turns to oversight, 67 percent have little or no confidence in the US government to regulate AI effectively, and 59 percent lack confidence that the companies building these systems will do so responsibly. The picture is not a public that hates the technology so much as one that distrusts the institutions steering it.

The data-security worry deserves its own line because it is the most universal. At 71 percent, the share of Americans who expect AI to make their personal information less secure cuts across age and gender and every level of usage alike, in a way few other findings do. It is one thing to feel abstract dread about society and another to believe a specific, personal harm is coming for your own data. That concern is also the hardest for the industry to wave away, since it maps onto a steady drumbeat of real breaches and opaque data practices that predate the current AI wave. When people say they distrust AI, a large part of what they mean is that they expect it to leak or misuse what it knows about them.

Views on the self are slightly warmer than views on society. Pew recorded 23 percent expecting a positive personal impact against 31 percent expecting a negative one, a narrower split than the lopsided societal numbers. People seem readier to imagine AI helping them individually than helping the country as a whole, which fits a long pattern in technology polling where the abstract future looks scarier than the concrete present.

The expert and public gap in AI public sentiment

One number does more than any other to explain the unease, and it is the distance between the people who study AI and everyone else. Pew has compared a nationally representative sample with a panel of 1,013 AI experts, and the split is stark. Among experts, 56 percent expect AI to have a very or somewhat positive impact on the United States over the next 20 years. Among the general public, that figure is 17 percent. The optimists and the worried are not looking at different facts so much as standing on opposite sides of a trust line.

That gap is the part that should give the industry pause. When the people closest to a technology are wildly more hopeful than the people who will live with it, the usual explanation is either that the experts know something the public has not yet grasped, or that the experts are insulated from the costs the public fears. Both can be true at once. Researchers see the capability curve and the upside, while a worker reading about layoffs or a parent worried about misinformation weighs a different ledger. Pew's own data shows only about 23 percent of Americans think AI will have positive effects on how people do their jobs, which is exactly the lived concern that no benchmark score addresses.

Concern has not stopped adoption

Here is the contradiction that makes the moment interesting. The same public that distrusts AI is using it in growing numbers. Pew found that 49 percent of Americans now use AI chatbots, up sharply from 33 percent in 2024, and 24 percent use one daily. ChatGPT leads that usage at 44 percent, a share that has roughly doubled since 2023. Around 60 percent of adults regularly read the AI-generated summaries that now sit atop search results, often without going out of their way to find them.

The age breakdown sharpens the irony. Younger adults use these tools the most and trust them the least. Among people under 30, some 57 percent use ChatGPT, yet 48 percent of that same young cohort predict a negative societal impact, the most pessimistic reading of any age group. Familiarity, in this case, is not breeding comfort. The people who reach for a chatbot every day are also the ones most convinced it is steering society somewhere bad, which suggests their worry comes from understanding the tools rather than fearing the unknown.

The usage data also reveals a market that is far less monolithic than the headlines suggest. ChatGPT may dominate at 44 percent, but Pew's reporting shows Gemini reaching 24 percent of users, Copilot 17 percent, Meta AI 14 percent, Grok 8 percent, and Claude 6 percent. Many people now touch several assistants depending on the task, which means the public is not forming its opinion of AI from one product but from a whole shifting landscape of them. That breadth of exposure cuts against any notion that distrust is rooted in unfamiliarity. A large share of Americans have tried multiple systems and still landed on caution.

Gender and age add further texture. Men report using AI more heavily than women, with roughly 27 percent of men using it daily against about 20 percent of women, and men tend to express more enthusiasm. At the other end, nearly three quarters of adults aged 65 and older say they never use AI chatbots at all. The result is an opinion split along familiar lines of access and exposure, where the heaviest users skew younger and male, and the most disengaged skew older. Yet even among the engaged, approval stays low, which is the detail that keeps undercutting the industry's hope that adoption will eventually soften the mood.

Why AI in branding became a turnoff

If the Pew survey measures how people feel about AI in the abstract, a second report measures what happens when companies wave it in their faces. WordPress VIP surveyed 1,200 US consumers for its Future of the Web 2026 research and found that 60 percent view the word AI in brand communications as unappealing rather than a selling point. Slapping AI on a product, it turns out, now repels more customers than it attracts.

The supporting numbers explain why. The study reported that 61 percent of consumers could not name a single brand using AI messaging well, and 16 percent believe no company is using it effectively at all. That is a marketing problem dressed as a feature. When a label inspires a shrug or a wince instead of curiosity, it has stopped doing the job a label is meant to do. The report also found that 74 percent of consumers see the internet as less human than it was a decade ago, a backdrop of fatigue that makes an AI badge feel like a warning rather than a promise.

That line, quoted in the WordPress VIP research, captures the disconnect cleanly. Companies have raced to advertise AI because it signals innovation to investors and the tech press, the very audiences Pew shows are most optimistic. The customer paying the bill is a different person entirely, one who wants a problem solved and does not care whether a model or a human solves it. To that buyer, an AI label is at best irrelevant and at worst a signal that corners may have been cut on the human touch they actually wanted. The study even put a rough clock on the friction, citing an average of about 40 minutes of automated interaction before consumers start to feel they are dealing with something synthetic and lose patience.

The same research exposes how much effort firms are pouring into a problem the public never asked them to solve. WordPress VIP reported that teams spend an average of 16.6 hours a week trying to improve their AI visibility, chasing placement in chatbot answers and AI summaries, while the measurement tools for that work remain fragmented across categories that range in price from free to six figures. The takeaway is a little bleak of irony. Companies are investing serious time and budget to be seen by AI systems, even as their human customers grow more allergic to being marketed to with the technology. The report also noted that no brand has yet clearly won at AI visibility, which leaves the whole scramble looking like motion without a finish line.

The trust deficit beneath both reports

Lay the two studies side by side and the same fault line runs through both. The Pew finding that 59 percent of Americans distrust AI companies to act responsibly is the wholesale version of the WordPress VIP finding that 60 percent recoil from AI in marketing. One measures belief about the industry, the other measures reaction to its sales pitch, and they point in the same direction. The public is not rejecting the underlying capability, which it uses daily, so much as the institutions and the hype wrapped around it.

This reading is supported by where the optimism survives. Personal-use numbers are warmer than societal ones, individual benefit polls higher than collective benefit, and actual usage keeps climbing even as approval sinks. Smart devices tell the same divided story. Pew found that 60 percent of adults regularly read AI-generated search summaries, while ownership of smart speakers sits at 35 percent and smartwatches at 37 percent, with smart doorbells trailing at 18 percent. People let AI into their homes and their search results while reporting, in the same breath, that they do not trust the companies behind it. People are voting with their hands while frowning with their mouths. The frustration attaches to the framing, the promises, the regulatory vacuum, and the sense that decisions are being made by parties they do not trust. Coverage at Gizmodo framed the Pew result in the same terms, as a public that has grown wary of the people promising the future more than the technology itself.

There is a historical rhyme here worth naming. New technologies often arrive to a chorus of optimism from their makers and suspicion from everyone else, and the suspicion usually softens as the benefits become tangible and the harms get regulated. What is unusual about this moment is the speed of adoption running ahead of any such softening. People are not waiting on the sidelines to be convinced. They have already woven AI into work and search and home life, and they have done so while their confidence in the companies behind it keeps falling. That ordering, deep use before earned trust, is the part that makes the current mood hard to predict, because the usual feedback loop where familiarity breeds acceptance does not seem to be firing.

What a skeptical public means for the year ahead

The temptation is to read 16 percent as a verdict on AI. It is better read as a verdict on AI's stewards. The technology is being adopted at a pace that any other product category would envy, with chatbot use leaping 16 points in two years, from 33 percent in 2024 to 49 percent in 2026, while the goodwill that usually rides alongside fast adoption has gone missing. That combination is unusual and unstable. Companies are winning the usage war and losing the trust war at the same time, and the WordPress VIP data suggests the trust deficit is now concrete enough to dent marketing.

For anyone building or selling AI products, the two reports carry a consistent and inconvenient message. Leading with the technology as the selling point is increasingly a mistake, because the audience that finds the word exciting is small and shrinking outside investor circles. The audience that pays wants outcomes and privacy, plus a sense that someone responsible is in charge, exactly the things Pew shows are in short supply. Closing the expert-public gap will not come from better demos or louder claims about capability. It will come from regulation people believe in and from companies behaving in ways that earn back the 59 percent who doubt them, alongside products that quietly work without demanding applause for the machinery inside. It is also worth noting what the surveys do not show. They do not show a public that has turned off AI, abandoned the tools, or organized against them. Usage is up, not down. What has eroded is the benefit of the doubt, the assumption that the people in charge have the public interest in mind. That is a softer thing to lose than market share, but it is also harder to win back, because it is built on a record rather than a pitch. A company can ship a better model in months. Earning back the trust of the 59 percent who doubt it will take far longer, and no amount of capability will substitute for it.

Until then, AI public sentiment is likely to stay split in this strange way, with the hands reaching for the tools and the minds bracing for the worst, and the gap between Wall Street's confidence and the public's wariness shows no sign of narrowing on its own.