Americans are picking up AI chatbots fast, and souring on them just as fast. According to a new Pew Research poll covered by The Verge AI, 49 percent of Americans now use chatbots at least occasionally, but 63 percent think the technology is moving too quickly. That gap between use and trust is the real story here.
The adoption curve is steep. Pew found chatbot use jumped from 33 percent in 2024 to 49 percent now. ChatGPT alone has doubled its reach since 2023, with 44 percent of respondents saying they’ve tried it. People are clearly finding reasons to open these tools.
They’re just not feeling good about it. Only 16 percent of Americans believe AI will have a positive impact on society. Sit with that. Half the country uses the tech, and barely one in six thinks it’s good for us.
The young are the harshest critics
Here’s the part that flips the usual script. You’d expect younger users to be the optimists. They’re not.
The Verge AI reports that 66 percent of people aged 18 to 29 use chatbots, the highest of any group. Yet 48 percent of that same group think AI will have a negative impact on society, and only 14 percent see it as positive. The more they use it, the less they seem to like where it’s headed.
Older generations flip the pattern: less usage, but softer opinions. Familiarity, in this case, breeds skepticism.
A quick read of the numbers:
- 49% of Americans use AI chatbots at least occasionally (up from 33% in 2024)
- 44% have used ChatGPT specifically (double the 2023 figure)
- 63% think AI is advancing too quickly
- 16% believe AI will benefit society
- 66% of 18-to-29-year-olds use chatbots, but 48% expect a negative impact
Where the real usage lives
The youngest group uses AI most broadly, but the heaviest daily users sit in the 30-to-49 bracket. Pew found 34 percent of them reach for a chatbot once a day or more.
Work is driving a lot of that. Roughly four in ten Americans said they’ve used AI for work tasks. Thirty percent think it makes them more productive, and 28 percent say it keeps them better informed.
That last claim deserves a flag. The same Pew research from 2024 found 66 percent of US adults were worried about AI spreading inaccurate information. So a chunk of people leaning on AI to stay informed are also the ones who don’t trust what it tells them. That’s a tension worth watching.
What this means for you
If you’re building or selling anything AI-powered, the takeaway is sharp. Usage is not the same as trust, and right now you can have a wall of users who quietly think your category is bad for the world.
A few practical reads:
- Adoption is no longer the hard part. Convincing people the tech is a net positive is.
- Your youngest, most engaged users may also be your loudest skeptics. Build for that, don’t ignore it.
- Accuracy is the trust lever. With two-thirds of adults worried about bad information, anything you ship that gets facts wrong compounds the problem for everyone.
- “AI makes you more productive” lands with about a third of people. The other two-thirds need a different pitch.
One limitation to keep in mind: this is survey data on attitudes, not behavior. People’s stated fears and their daily clicks often diverge, and the rising usage numbers prove it. Sentiment can shift quickly as the tools improve or stumble.
The bigger picture is a public that’s adopting AI faster than it’s making peace with it. That’s an unstable spot, and it won’t hold forever. The companies that close the trust gap, not just the usage gap, are the ones that win the next round. You can find the full breakdown over at The Verge AI.