The kids are not begging for AI, and that breaks the demand story

On August 13, 2026, MIT Technology Review published a set of interviews with kids aged 10 to 18 about AI, reported by Jen Swetzoff and Keeley McNamara, the founding editors of Anyway, an independent print magazine for tweens and teens ^[1][2]. The line in it that matters for anyone building consumer AI: these kids are not begging for it, the way they beg for iPhones and Snapchat accounts ^[3].

The interviewers expected stories about cheating, inspiring uses, and worries about deepfakes and job destruction ^[4]. What they got instead included "bruh" and "meh," and kids so opposed to AI or so uninterested in it that they did not want to discuss it ^[5]. One teen said she will not touch AI because of the environmental impact ^[6]. Winter, 17, told them "AI isn't the solution to our problems" and "I'm afraid it's going to be the end of creativity and critical thinking" ^[7]. Most of the kids interviewed did admit to using AI at least a little ^[8], which is the point: usage without wanting.

The distribution channel is the tell. Many said their first encounters with AI came from their parents or their schools, or that it was simply embedded in devices and apps they already used, such as a Google search ^[9]. That is institutional push and default placement, not the peer-to-peer pull that carried Snapchat. Products acquired that way do not get the benefit of the doubt when they are bad, and they do not generate the organic advocacy that consumer growth models assume.

The survey data lines up. Pew Research Center, in work published in February 2026, found 57% of US teens had used chatbots to search for information, 54% for help with schoolwork, and 47% for fun or entertainment, but only 12% for emotional support or advice ^[10]. Information search runs about 4.8 times the emotional-support use ^[11]. If your roadmap assumes teenagers are the leading edge of AI companionship, that 12% is the number to argue with.

The second-order costs land on kids without their consent. Remy, 16, in New York said the word he associates with AI is "indifferent," and that he uses the free version of Claude for programming outside school, including a program to tweak his computer's overclock ^[12][13]. He also said AI mostly makes his school assignments worse: his English class moved to in-class writing because teachers do not want cheating, leaving 70 minutes to produce a whole essay, which he thinks hinders his writing ^[14]. He noted that Princeton voted to let faculty proctor exams for the first time in over a century, ending an honor code dating to 1893 ^[15], roughly 133 years of it ^[16]. The product shipped; the assessment regime degraded around it.

Treat the sentiment carefully. These are edited interviews, not a representative sample ^[17], and the reporters' own read is that teens are over four times more likely to use AI in innocuous ways than potentially harmful ones ^[18].

Watch whether Pew's next wave moves the 12% emotional-support figure, since that is where the companion-product theses live. Watch school and district procurement, because that, plus defaults inside existing apps, is currently doing the acquisition work that youth demand is not.