Classrooms across the United States now hum with a new, largely invisible presence. Artificial intelligence is open on laptops, active on smartphones, and moulding how teenagers complete assignments, search for answers, and even think about their futures.A survey by the Pew Research Center, conducted between September 25 and October 9, 2025, among US teens aged 13 to 17, captures the scale of this shift. Sixty-four percent of teens say they have used AI chatbots, and about three in ten report doing so daily. Tools such as ChatGPT, Copilot, and Character.ai have moved swiftly from novelty to routine.
Homework help becomes mainstream
Academic support sits at the heart of teenage AI use. Pew reports that 57% of teens have used chatbots to search for information, while 54% have used them for help with schoolwork. Entertainment is part of the story, 47% say they use chatbots for fun, but study-related purposes clearly dominate.Patterns of school use reveal something more layered than simple shortcut-seeking. One in ten teens say chatbots help with all or most of their schoolwork. Another 21% say AI supports “some” of their work, and 23% say it helps with “a little.” Forty-five percent report not using chatbots for school assignments at all.Research assistance and mathematics top the list of academic tasks. About four in ten teens say they have used chatbots to research a topic or solve math problems. By comparison, 35% say they have used AI to edit something they wrote. Many teenagers appear to treat chatbots as on-demand explainers and problem-solvers rather than silent essay substitutes.Helpfulness ratings are strikingly high. Roughly a quarter of teens describe chatbots as “extremely” or “very” helpful for completing schoolwork, and another 25% call them “somewhat” helpful. Only 3% say these tools were of little or no help. For schools still debating policy boundaries, that endorsement complicates enforcement.
Cheating, perception, and pressure
Academic integrity remains a flashpoint. Fifty-nine percent of teens say students at their school use AI chatbots to cheat at least somewhat often. About a third believe this happens extremely or very often. Fourteen percent say it rarely or never occurs, and 15% are unsure.Teens who have used chatbots for schoolwork are especially likely to perceive cheating as common: 76% of this group say it happens at least sometimes at their school. That perception alone reshapes classroom culture. When students believe AI-assisted cheating is widespread, the ethical ground shifts, even for those who choose not to participate.Teachers face a delicate balancing act. Artificial intelligence can function as a tutor and accessibility tool, yet it also blurs authorship and accountability. Policy conversations often lag behind the pace of student experimentation.
Personal conversations with machines
Concerns about emotional reliance surface in public debate, and the data offer a measured perspective. Sixteen percent of teens say they have used chatbots for casual conversations. Twelve percent report seeking emotional support or advice. Majorities say they have not engaged with chatbots in these personal ways.Even minority usage carries weight. Adolescence is marked by vulnerability and self-discovery. When some teens turn to algorithms for counsel, questions arise about how digital companionship intersects with human relationships.
Optimism for themselves, caution for society
Looking ahead 20 years, teens express guarded optimism about AI’s personal impact. Thirty-six percent believe it will affect their own lives positively, while 15% expect a negative personal effect. Uncertainty is notable as well, with 17% unsure.Views grow more critical when society enters the frame. Twenty-six percent think AI will have a negative impact on society, compared with 31% who expect a positive one. Eight percent remain unsure.Open-ended responses collected by Pew research reveal the reasoning beneath those numbers. Teens who foresee benefits often mention efficiency, productivity and faster access to information. Some argue that AI will automate repetitive tasks and create room for creativity. Others stress that learning to use AI will be essential for future success.Those predicting harm frequently cite overreliance and the erosion of critical thinking. Among teens who expect a negative societal impact, 34% mention dependence or loss of creativity. A quarter raises fears about job loss. Roughly one in ten point to misinformation or difficulty distinguishing authentic content from AI-generated material, and an identical share mentions misuse or broader threats.Their concerns mirror debates playing out among policymakers, business leaders, and researchers, suggesting that teenagers are attuned to the broader stakes.
Human judgment still holds weight
Confidence in AI’s capabilities remains measured. Roughly half of teens say today’s AI would do worse than humans at making hiring decisions, while only 15% believe it would do better. Many also think AI would underperform humans in driving, making medical diagnoses and writing songs.Opinions are more divided on customer service. Teaching a skill stands out as the only task where a larger share of teens, about a third, believe AI would outperform humans, compared with roughly a quarter who think it would do worse. Substantial portions, often around a quarter, say they are unsure how AI would perform in complex areas such as health care or employment.
Familiarity without full confidence
Awareness of AI chatbots is nearly universal. Pew research finds that 56% of teens say they have heard “a lot” about them and 39% say they have heard “a little.” Just 5% report hearing nothing at all.Confidence levels vary more widely. About a quarter of teens describe themselves as extremely or very confident using chatbots. Roughly three in ten say they are somewhat confident. Around one in ten acknowledge little to no confidence.Exposure has arrived faster than fluency. Many teenagers are comfortable navigating AI tools, yet not all feel fully in control of them.
A generation negotiating its tools
Teenagers across the United States are not passive recipients of artificial intelligence. They are active participants in shaping how it fits into their academic routines, social interactions and future aspirations. They see opportunity in its speed and adaptability, yet they recognize risks tied to dependence, employment and truth itself.Artificial intelligence has settled into daily teenage life with remarkable speed. Whether that integration strengthens critical thinking or gradually erodes it will depend less on the technology alone and more on how schools, families and students themselves choose to engage with it.
