AI Pisses Off Gen Z. They Use It Anyway.

22% say they're excited about AI. 31% say they're angry. 51% use it every week or every day. Same study, same generation, same technology, released on April 9, 2026.

The source: Gallup's Voices of Gen Z survey, conducted on 1,572 Americans aged 14 to 29, fieldwork between late February and early March 2026. The picture isn't a rejection of AI. It's something more uncomfortable. Steady, even massive use, paired with a collapse in positive sentiment.

Excitement dropped 14 points in a year. Anger climbed 9. Hope lost 9. Meanwhile, the usage curves haven't moved.

Adoption holds steady, sentiment cracks

For three years, the consumer AI market has been steered by a single metric: the adoption rate. It's the number VCs put in their decks, that companies dust off in board meetings, that media outlets lead with. "X% of Americans use ChatGPT every month." It became a compass.

The problem with a compass: it tells you a direction. It doesn't tell you whether you actually want to go there.

The Gallup data shows those two things are pulling apart. Weekly AI use among American Gen Z has been flat since March 2025. Over the same period, excitement lost almost half its value, sliding from 36% to 22%. Last year's positive majority is this year's minority. And yet the same kids open ChatGPT, Gemini and Claude every single day.

The smartphone offered the playbook starting around 2015. Near-total penetration among young adults, hours of daily use, and in parallel a growing scientific literature on anxiety, sleep and attention.

Everyone used it, plenty hated it, and the industry kept publishing adoption charts as if everything was fine. Sentiment took years to become visible. By the time it did, it was too late to fix the product.

What Gen Z sees that no one else does

On the concrete questions, Gen Z is clear-eyed in a way the rest of the market should worry about. 80% think AI will hurt their learning. 48% of employed Gen Zers say AI's workplace risks outweigh the benefits. Trust in work produced by a human alone tops out at 69%. Trust in AI-assisted work crashes to 28%.

41-point gap between the two. Same question, same people, same task.

Zach Hrynowski, the senior Gallup researcher who led the study, flagged a detail that deserves attention. The angriest Gen Zers aren't the teenagers using ChatGPT to fudge their homework. They're the older ones, the ones leaving school and looking for their first real job. The young adults are seeing something else.

They see entry-level openings shrinking because AI does the intern's work for free. They see their degree quietly devaluing while still being the price of admission. They watch a labor market reshape itself underneath them, in real time, driven by a technology they barely had time to discover.

And they keep using the tool that's doing it, because there's no opt-out. The teacher grades the ChatGPT essay sitting next to theirs. The manager expects "AI-assisted" deliverables because the pace demands it. The friend group treats the abstainer as a Luddite.

It's a forced cohabitation with a tool people learned to tolerate, not enjoy. The adoption metric measures the cohabitation. It doesn't measure the lease.

The European mirror is already there

Gallup is an American study, but the angle doesn't stop at the Atlantic. Ifop, with Jedha AI School, ran a fall 2025 poll on 1,000 French respondents aged 16 to 25. The result is an almost perfect overlay.

89% have already used a generative AI tool. 73% use one at least weekly. 25% daily. Six in ten worry their job will disappear because of AI. And 85% want mandatory ethics and social-impact modules in any AI training curriculum, which isn't the posture of a generation comfortable with the deal it's been offered.

US side, French side, the paradox is the same shape. Record use, deteriorating sentiment, explicit fear about professional futures. The curves aren't crossing, they're moving apart. And in 2026, they're moving apart faster than they were in 2025.

What it telegraphs is probably a cultural backlash before a political one. The first signals won't come from Congress. They'll come from teachers refusing AI-generated work, from artists pushing anti-AI charters, from journalists who've stopped calling a language model "intelligent."

Gen Z is the first cohort to live full-time in this world. They're also the first to put numbers on the fatigue, in a probability survey funded by a foundation no one would accuse of hating tech.

For anyone steering AI products, content or strategy, that changes the reading frame. The adoption dashboard has effectively run its course. It measured the social pressure to use the tool. Not the desire to use it. And not how much patience is left before that pressure becomes politically untenable.