What kids actually think about AI (it's not what adults expected)
A new survey and dozens of conversations with 10-to-18-year-olds reveal a generation that is cautious, opinionated, and far less dazzled by artificial intelligence than the headlines suggest.
Key points
- A February 2026 Pew Research Center survey found that 57% of US teens had used AI chatbots, the text-based tools behind products like ChatGPT, to search for information.
- The same survey found only 12% of teens used chatbots for emotional support, making "therapy bot" use four times less common than ordinary information searches.
- Most young people in the conversations report using AI occasionally but deliberately, choosing what tasks to hand over and what to keep for themselves.
- At least one 18-year-old has built a pipeline of AI agents, software programs that carry out tasks automatically, to send plain-language government policy updates to voters.
- Environmental concerns, not job fears, are the worry that came up most often among teens who distrust AI.
Ask a teenager what they think about artificial intelligence and you might get a shrug. That was one of the findings MIT Technology Review gathered while interviewing kids aged 10 to 18: many gave responses like "meh" or simply did not want to discuss it. Some refused to engage with AI at all. Others used it every day.
The reality, it turns out, is messier and more interesting than either the panic or the hype.
How much are kids actually using AI?
Most teens use it at least a little, but rarely in the ways adults fear most. The Pew Research Center published figures in February 2026 showing that 57% of US teens had used a chatbot to look something up, 54% had used one for schoolwork, and 47% had used one for fun. Only 12% had turned to a chatbot for emotional support or advice.
| Use case | Share of US teens (Feb 2026) |
|---|---|
| Searching for information | 57% |
| Help with schoolwork | 54% |
| Fun or entertainment | 47% |
| Emotional support or advice | 12% |
For younger children, primary-school and middle-school age, AI barely registers as a want. Several kids said their first exposure came from a parent or a teacher, or they simply noticed it baked into apps they already used.
What worries them?
Job losses barely came up. What came up instead were creativity, critical thinking, and the environment.
"AI isn't the solution to our problems," said Winter, 17. "I'm afraid it's going to be the end of creativity and critical thinking." Hazel, 17, a rock climber who wants to become an ecologist, pointed to the water consumption of the data centres, the vast warehouses of computer servers that run AI systems. She stopped using the tools once she looked into the environmental cost.
Remy, 16, a programmer from New York, described AI as "sort of like the first car or the first airplane: interesting but crude." He uses Claude, Anthropic's chatbot, occasionally to help with personal coding projects. At school, he prefers writing his own code because AI-generated code, he says, is sloppy and hard to fix. He is more excited about AI flagging breast cancer in medical scans than he is about any consumer chatbot.
What does a teen building with AI actually look like?
Danielle, 18, from California, built a project called Next Voters with a team of peers. The system uses a chain of AI agents, each one a small program with a specific job, to find official government documents, verify them, sort them by topic, and deliver a plain-English summary by email each week. The goal is to make local policy accessible to people who will not wade through hundred-page meeting transcripts themselves.
She is clear about the boundary: "We don't want AI to think for people; we want to use it to disperse knowledge."
That boundary, knowing what to hand over and what to keep, is what most of the young people in these conversations kept coming back to. They are not against the technology. They are thinking carefully about which parts of their lives they want it touching.
For parents and teachers, the practical takeaway is straightforward. The conversation about AI is already happening, whether adults start it or not. Talking openly about what the tools can and cannot do, where they go wrong, and why the source of an answer matters is more useful than a blanket ban.
Common questions
Is AI doing kids' homework for them?
Some teens admitted their peers use it to write papers, but the Pew data suggests this is far less common than ordinary searching or entertainment use. Many schools have shifted to in-class writing to reduce the temptation.
Should parents be worried about kids using AI chatbots for advice?
Only 12% of teens in the February 2026 Pew survey used chatbots for emotional support. That is worth monitoring, since chatbots can give wrong or harmful answers, but it is not the dominant use pattern the data shows.
