FOLLOW US

Artificial intelligence has quietly woven itself into the daily routines of millions of Americans. People use it to search for information, finish work assignments faster, draft emails, and even process their emotions after a hard day. But here is the uncomfortable truth that a sweeping new Pew Research Center study has just laid bare: the more Americans use AI, the less they seem to trust it.

This paradox sits at the heart of a landmark survey conducted between February 17 and 23, 2026, involving 5,119 U.S. adults. The findings paint a vivid and somewhat unsettling portrait of a nation that has embraced a powerful technology it doesn't fully understand, doesn't entirely believe in, and increasingly fears is moving faster than anyone can control.

AI Adoption Is Surging, and the Numbers Are Striking

Let's start with the raw adoption numbers, because they tell a story of explosive growth.

According to Pew's data, roughly half of all U.S. adults (49%) now say they use AI chatbots. That is a dramatic leap from just one year earlier, when only about a third of Americans reported using these tools. In other words, AI chatbot usage has grown by roughly 50% in a single year, a pace that would be considered extraordinary in virtually any other technology category.

Even more telling is how frequently people are using these tools. A full quarter of all American adults say they use AI chatbots on a daily basis. Within that group, 12% say they use them several times a day, and a notable 4% describe themselves as "almost constantly" engaged with AI tools. That last figure, while small in percentage terms, represents millions of people whose relationship with AI has moved well beyond casual curiosity and into something approaching dependency.

This is no longer a niche technology for Silicon Valley insiders or tech-savvy college students. AI chatbots have crossed a critical adoption threshold and are now a mainstream American habit, sitting alongside social media, streaming platforms, and online shopping as fixtures of everyday digital life.

ChatGPT Still Dominates, But the Landscape Is Shifting

When it comes to which AI tools Americans are actually using, the pecking order is fairly clear.

ChatGPT, developed by OpenAI, remains the undisputed heavyweight champion of public AI adoption. According to Pew, 44% of U.S. adults report using ChatGPT, which is more than double the share from 2023. That kind of growth in just a few years is remarkable by any measure and confirms ChatGPT's grip on the popular imagination as the face of the AI revolution.

Google's Gemini comes in second place, a result that likely reflects the enormous advantage Google has in inserting its AI tools directly into products and services that hundreds of millions of Americans already use every day. Microsoft Copilot follows at 17%, benefiting from its integration into the Microsoft 365 suite that dominates corporate and educational environments. Meta AI registers at 14%, and Elon Musk's Grok sits at 8%.

One figure in particular stands out and deserves careful attention. Anthropic's Claude, which has generated enormous buzz in AI industry circles and has reportedly generated more revenue per user than ChatGPT according to recent reporting, barely registers at just 6% in the general public. This gap between industry reputation and public awareness is significant. It suggests that Claude's current strength lies primarily with professionals, developers, and power users rather than with the broader American population, and represents both a challenge and a major opportunity for Anthropic.

What Are Americans Actually Using AI For?

The survey digs into the specific use cases driving AI adoption, and the results offer a window into how deeply this technology has already penetrated different corners of American life.

Information searching tops the list, with 42% of U.S. adults saying they use AI chatbots to look things up. This is perhaps unsurprising given that AI chatbots were quickly positioned as a smarter, more conversational alternative to traditional search engines. For many users, asking an AI a question has simply replaced typing a query into Google.

Work-related tasks come in second, with 38% of employed adults reporting that they use AI chatbots to help them with professional responsibilities. This ranges from drafting documents and summarizing meetings to writing code, analyzing data, and generating marketing copy. The integration of AI into the workplace is accelerating rapidly, and these numbers likely understate the true extent of professional AI use given how quickly corporate adoption has been growing.

But perhaps the most striking and thought-provoking finding in this entire survey involves emotional support. One in ten Americans now report using AI chatbots for emotional support. Read that again: roughly 33 million Americans are turning to an artificial intelligence when they need someone, or something, to talk to about what they are going through emotionally.

Among adults under 30, the numbers are even more striking. One in five young adults say they use AI chatbots for emotional support. And 7% of that age group say they use AI for "companionship." These are not just statistics about technology adoption. They are data points about loneliness, mental health, and the way young people are redefining what connection looks like in the digital age.

The Paradox at the Center of It All: Heavy Users Are the Most Worried

Here is where the Pew research becomes genuinely fascinating, because it reveals a deep and striking contradiction at the heart of American AI adoption.

The people using AI the most are also the people who are the most worried about it.

Adults under 30 are the heaviest chatbot users, with 66% reporting that they use these tools. But they are also the most pessimistic about AI's long-term impact on society. A full 48% of adults under 30 predict that AI will have a negative impact on society over the next 20 years. Compare that to just 30% of adults over 50 who feel the same way.

Think about what that means. The generation that has most enthusiastically adopted AI, that uses it daily for work and personal tasks and emotional support, is simultaneously the generation most convinced that this technology will ultimately make the world worse. They are using something they believe may be harmful. That is not the behavior of a generation that has been seduced by technology into uncritical acceptance. It is the behavior of a generation that feels it has no choice but to adopt tools it has deep reservations about, because not adopting them means falling behind in school, at work, and in life.

Overall, the numbers are sobering for AI optimists. Forty percent of all Americans predict that AI's impact on society will be negative. That figure is nearly triple the 16% who see a positive outcome. About 31% expect the impact to be roughly equal in terms of upsides and downsides. In other words, when you add up the pessimists and the ambivalent, the AI industry is looking at an American public that is, at best, deeply uncertain about where all of this is heading.

Speed Kills Trust: Why Americans Think AI Is Moving Dangerously Fast

If there is one theme that cuts across every demographic in Pew's study, it is the belief that AI is moving too quickly.

A striking 63% of Americans say AI is advancing "too fast." Just 2% say it is progressing too slowly. That is a thirty-to-one ratio of concern to enthusiasm about the pace of development. Whatever excitement may exist about AI's potential, it is being drowned out by anxiety about whether anyone is in control of where this technology is going.

The fear that stands above all others is data security. Seventy-one percent of Americans believe that AI will make their personal information less secure. Only 3% think AI will improve data safety. This is not a minor concern about edge cases. It is a majority view that one of the most personally significant consequences of AI's rise will be a degradation of individual privacy and data protection.

Then there is the question of oversight and governance. Who is supposed to be making sure all of this goes well? According to Americans, nobody is doing a particularly good job of it.

Sixty-seven percent of those surveyed said they have little or no confidence in the U.S. government to regulate AI effectively. That figure is up from 62% in 2024, suggesting that confidence is eroding rather than growing. Nearly 59% said they lack confidence in American companies to develop and use AI responsibly. These are devastating numbers for an industry that has repeatedly promised to take safety and responsibility seriously.

There is also a fascinating partisan shift buried in these numbers. Democrats have grown dramatically more skeptical of the government's ability to regulate AI over the past two years. The share of Democrats who lack confidence in government regulation has jumped from 54% to 74%, a 20-point swing. Meanwhile, Republicans have actually become somewhat less concerned, dropping from 70% to 61%. This reversal likely reflects broader political dynamics, including which party currently controls the government and how different constituencies feel about federal regulatory authority in general.

AI Is Already Everywhere, Even Where People Don't Realize It

One of the most eye-opening sections of the Pew report deals with how thoroughly AI has already embedded itself in American consumer technology.

Thirty-seven percent of U.S. adults own a smartwatch. Thirty-five percent have a smart speaker such as an Amazon Echo or Apple HomePod. Eighteen percent have a smart doorbell equipped with AI capabilities. Thirteen percent own a robot vacuum with AI. And 11% have a smart thermostat powered by AI.

But the most revealing number may be this: 60% of Americans say they have encountered AI-generated summaries in their search results. More than half the country is already interacting with AI on a regular basis without necessarily thinking of it as "using AI." The technology has become so embedded in the fabric of everyday digital life that many people are consuming AI output without ever consciously making a decision to engage with it.

This has important implications for how we think about AI adoption numbers. The 49% who say they use AI chatbots may significantly undercount the total proportion of Americans who are actually interacting with AI in some meaningful way every day.

The Racial and Gender Divide in AI Adoption

Pew's data also reveals important differences in AI adoption and attitudes across racial and gender lines.

Among racial groups, Asian adults show the highest rate of chatbot adoption at 70%, making them the most enthusiastic AI users of any demographic group. They are also the only group where positive views about AI's impact outnumber negative ones. This is a significant outlier in an otherwise largely pessimistic landscape.

Gender differences in adoption are smaller than might have been expected. Men and women now use chatbots at roughly equal rates: 50% for men versus 47% for women. However, men are more likely to be daily users, with 27% of men using AI chatbots every day compared to 20% of women.

On the question of attitudes, women tend to be more pessimistic about AI's personal impact than men. Thirty-three percent of women expect AI to have a negative impact on them personally, compared to 27% of men. This gender gap in AI skepticism is a relatively consistent finding across research on technology attitudes and may reflect concerns about how AI is being designed, who it is being designed for, and what the economic consequences of automation will be for industries where women are heavily represented.

The 51% Who Refuse to Use It: What Non-Users Are Saying

While adoption is growing, a majority of Americans, 51%, still say they don't use AI chatbots at all. Understanding why matters enormously for the future trajectory of AI adoption.

Among non-users, 60% cite lack of interest as a major reason for avoiding these tools. They simply don't see a compelling reason to use them. Fifty-four percent raise privacy concerns, reflecting the broader data security anxiety that runs through the entire survey. Forty-five percent say they don't trust the accuracy of AI-generated information, a concern that has been validated repeatedly by high-profile cases of AI systems producing confident-sounding but factually wrong answers. Twenty-nine percent say they don't know how to use AI chatbots, suggesting that ease of use and digital literacy remain meaningful barriers.

Only 3% of non-users say social judgment is a reason they avoid chatbots, which tells us something interesting: the stigma around AI use has essentially evaporated. Not using AI is no longer embarrassing; it is simply a personal choice that most non-users feel perfectly comfortable making without worrying about what others think.

Most non-users also appear settled in their decision. Sixty-seven percent say they are unlikely to use a chatbot in the next 12 months. That is not a population that is on the fence. That is a population that has made a considered decision to opt out, at least for now.

What This All Means for the Future of AI

Step back and look at everything Pew has found, and a coherent picture emerges, even if it is an uncomfortable one for the AI industry.

Americans are adopting AI at extraordinary speed. The tools have become cheap, accessible, and genuinely useful for a wide range of tasks. Young people especially have incorporated AI into their work, their creative lives, and even their emotional lives in ways that would have seemed implausible just five years ago.

But adoption is not the same as acceptance. It is not the same as trust. Millions of Americans are using AI because it is useful, or because they feel they have to in order to stay competitive, not because they have been persuaded that AI's long-term impact will be good. They are going along with a technological transformation they didn't fully choose and that they largely believe is moving too fast, protecting their data poorly, and operating without adequate oversight.

The AI industry has spent enormous energy demonstrating what its products can do. The results speak for themselves in terms of capability. But the harder challenge, and the one that Pew's data suggests is far from solved, is demonstrating that AI can be trusted: trusted with personal data, trusted to produce accurate information, trusted to operate within responsible boundaries, and trusted to develop in ways that serve broad human interests rather than narrow commercial ones.

Until that trust gap is addressed, the picture Pew has captured will persist: a nation of users who have decided to adopt a technology they don't fully believe in, can't entirely control, and quietly suspect may be leading somewhere they don't want to go.