What happened?

MIT Technology Review conducted a series of interviews with teens aged 10 to 18 to understand their attitudes toward AI. According to the outlet's report dated August 13, 2026, teens' reactions turned out to be far more complex than expected; some remained completely indifferent while others expressed deep concerns. Winter, 17, one of the interviewees, said she worries AI could mean the end of creativity and critical thinking.

According to the outlet, Pew Research Center's survey published in February 2026 found that 57% of US teens have used chatbots to look up information, 54% have used them for homework, and 47% have used them for entertainment. The share of those using AI for emotional support or advice was only 12%.

Why does it matter?

The interviews show that teens' approach to AI is far more nuanced than adults assume. Remy, 16, said AI use at school is restricted due to cheating concerns, and this has negatively affected their own writing skills. Remy also noted that Princeton University changed its honor code, which had been in place since 1893, due to AI, deciding to have proctors present during exams.

Danielle, 18, described using AI in a project called Next Voters. The project aims to increase teens' political engagement by summarizing local government documents in plain language. Some teens, like Hazel, 17, keep their distance from AI due to concerns about data centers' water consumption and environmental impact.

What we know

  • Most of the interviewed teens acknowledged using AI at least to a limited extent.
  • For elementary and middle school-aged children, AI is not yet as sought-after a technology as the iPhone or Snapchat.
  • Teens' first experiences with AI usually happen through their parents or schools.
  • Some teens use AI to develop tools such as characters, technology platforms, or learning aids.

What's next?

The report emphasizes that teens evaluate AI more in terms of concern about societal harm than fear of job loss. The authors argue that the danger is a reason to teach the tool rather than ban it, noting that teens want to retain control over how they use AI. No concrete timeline for new surveys or policy steps on the matter was shared.

What the survey and the interviews each show

The most striking figure is not the most-discussed use but the least: 12 percent of teens turn to AI for emotional support or advice. Public debate is built around adolescents forming emotional bonds with chatbots, while the data says that is a minority behaviour. Twelve percent is not a small number, but it is not large enough to sit at the centre of the conversation either.

The second point is method. Interviews are not a survey; a handful of teenagers in their own words, however well chosen, do not represent the age group. The value of reporting like this is not representativeness but surfacing reasons a questionnaire never captures: Hazel objecting on the grounds of data-centre water consumption is a form of refusal no multiple-choice option would have registered.

What Remy describes points at a similar gap. A ban imposed out of concern about cheating makes learning to write harder, not using the tool harder. The choice between banning and teaching is a problem these accounts make visible in a way survey percentages cannot.