AI was supposed to have won everyone over by now. It hasn't.

The trend no one in the industry wanted to see

You'd expect that, after years of AI tools becoming ever more powerful and widespread, people would be warming up to them. According to recent surveys, the opposite is happening. Pew research found that 52% of Americans now say they're more concerned than excited about AI's growing role in daily life — a sharp rise from 37% a few years earlier.

Other surveys tell a similar story: younger adults in particular say they don't trust the leaders of major AI companies to act responsibly, and a solid majority of Americans think AI is being developed and rolled out faster than is wise.

Why people are turning more negative

A few recurring themes stand out. First, AI features keep showing up uninvited in products people already use — email, TVs, search engines — often without anyone asking for them. Second, there's genuine unease in schools, where AI has made it much easier for students to cheat on assignments, from middle school through college. Third, there are ongoing conflicts over AI companies training their models on the work of artists, musicians, and writers without paying them or asking permission. And underlying all of this is a simple gap between promise and experience: for a lot of ordinary people, AI feels more like a threat to their job than something that has actually improved their daily life yet.

People are pushing back in small, visible ways

This discomfort is also starting to seep into everyday habits — sales of old-fashioned "dumb phones," cassette decks, and classic iPods are up, and hobbies with zero AI involvement, like quilting, knitting, and mahjong, are seeing a resurgence. More broadly, there's a noticeable shift back toward real-life activities, away from digital ones.

Even AI leaders admit it

Perhaps the most telling sign is that the people running these companies are saying it out loud themselves. One prominent tech CEO said the industry needs to "build products that ordinary people can actually use," rather than impressive demos. And Anthropic's own CEO described the moment as a "crisis of trust," openly admitting that the industry "has not yet delivered on its big promises to change the world for the better." When the people selling AI are the ones naming the trust gap, that's a fairly strong signal that the gap is real.