For today’s teens and tweens, artificial intelligence is neither a simple promise nor a simple threat. In conversations with kids aged 10 to 18, the clearest pattern was not blind enthusiasm or blanket rejection, but a cautious, uneven effort to decide where AI belongs in their lives.
Some described AI as helpful for school, coding, studying, medical routines or creative play. Others called it disheartening, environmentally troubling or simply boring. Even among kids who use it, the strongest message was control: they want tools that assist them, not systems that replace their thinking.
AI is already present, but not always wanted
Many of the younger people interviewed did not describe AI as something they had actively chased. It was often introduced by parents, schools or the apps and devices they already use. One example was search: AI can simply appear inside a familiar Google search experience.
That matters because it changes the question. Kids are not only deciding whether to download a new tool. They are also reacting to AI as something that has started appearing around them, sometimes before they have formed a strong opinion about it.
The responses ranged from indifference to concern. One teen said the word that came to mind was “indifferent,” explaining that school, tae kwon do, reading and games with friends did not need AI. Another, Winter, a 17-year-old, put the concern more sharply: “AI isn’t the solution to our problems,” and “I’m afraid it’s going to be the end of creativity and critical thinking.”
Still, most of the kids interviewed had used AI at least a little. Their accounts suggest that rejection and use can coexist. A teen can dislike the direction of AI and still try it for homework help, programming or a practical task.
School is where the tradeoffs become obvious
Education came up again and again. Some kids used AI to revise writing, debug code, generate study ideas or create memory tricks before a test. Others saw it changing school in ways they did not like.
One student said AI had made English assignments worse because teachers now require in-class writing to prevent cheating. That meant only 70 minutes to write a whole essay, which the student felt hindered writing. The same student pointed to Princeton voting to let faculty proctor exams for the first time in over a century, saying its honor code goes back to 1893.
Other students were trying to build tools around learning rather than avoid learning. One teen created an AI tutor called Aceflow after looking for a way to study less while keeping grades up. The tool could take a 30-minute lecture video on YouTube, a blog post, a PDF of a textbook or presentation slides, then create practice questions and a tutor that explained material in a teacher’s style.
The student built it for personal use, showed friends, posted it on TikTok and said it got tons of views on TikTok and thousands of users. The argument for the tool was not that AI should do schoolwork for students. It was that personalized help should not be limited to families who can hire expensive private tutors.
Young users see both practical help and real limits
The kids’ examples were often concrete. One used Claude, the free version, for programming outside school, including help writing a program to see whether a computer’s overclock could be tweaked. The same student had also tested AI for game development and found it weak, saying it produced sloppy code and struggled to blend mechanics into a cohesive result.
Another student used AI in daily life after being diagnosed with type 1 diabetes in January. ChatGPT helped with recipes by working out how many carbs were in a serving. A glucose monitor and insulin pump also communicated using their own kind of AI to predict dosing, with the monitor tracking blood sugar and the pump doing the math.
That system could give a correction dose if it predicted blood sugar would be high in 30 minutes, or stop insulin if it predicted a low was coming. The student said the pump could catch lows at night, the phone would go off, and then juice could help. The practical result was simple: more sleep.
But even that student named a limit. The hardest part of diabetes was not something AI could solve: remembering to carry supplies everywhere.
Creativity is a boundary many kids want to protect
Several kids worried less about AI taking jobs and more about what it could do to society, creativity and critical thinking. The concern was not only whether a chatbot could generate an answer. It was whether people might stop doing the hard parts themselves.
One young artist compared AI to a calculator: a calculator does math, and AI does other things. But the same student wanted drawing and songwriting to remain personal work. Another student used ChatGPT to catch grammar and spelling errors in a short story, but worried about AI robbing people of the ability to think creatively or think for themselves.
AI entertainment also came with caution. Character.AI was described as more fun than chatting with ChatGPT, because users can create characters and share them. The examples included joining the pirate crew from One Piece on Netflix through an existing character, plus games such as a rap-star simulator, World War II simulators and chats involving assassins from a TV show.
Yet the warning was direct. The content was described as pretty unfiltered, and the free version’s memory could run out, causing chats to slow down and forget what happened. The conclusion was practical: if AI is left to run on its own, it falls apart, so the user has to keep steering it.
The data matches the nuance
The interviews fit with a survey published in February 2026 by the Pew Research Center. It found that 57% of teens in the US had used chatbots to search for information, 54% had used them to help with schoolwork, and 47% had used them for fun or entertainment. Only 12% had used them for emotional support or advice.
That pattern matters. The most common uses described in the survey were information, school and entertainment, not emotional dependence. The source article notes that teens are over four times more likely to be using AI in innocuous ways than potentially harmful ones.
None of this removes the risks. Kids can run into unfiltered content, rely too heavily on a chatbot, or trust an answer that is wrong. But the interviews point toward a better response than pretending AI is not there: teach young users how to question it, verify it and decide when not to use it.
The strongest lesson from these kids is not that AI is good or bad. It is that many of them already understand the central problem. They can see the convenience, the flaws, the pressure and the temptation. And despite growing access to tools that can write, answer, suggest, summarize or simulate, many still want to keep their own hands on the work that matters.