Teens using AI

More than half (54%) of U.S. teens ages 13 to 17 have used AI chatbots for help with schoolwork, according to Pew Research Center, up from 26% using ChatGPT for schoolwork in 2024 and 13% in 2023. About 10% say they do all or most of their schoolwork with chatbot help, and 64% use chatbots for any purpose. Common Sense Media and College Board surveys add more high school detail below.

Pew Research: teens ages 13–17 (primary)

Findings from Pew Research Center — How Teens Use and View AI. Related trend: Pew Research Center.

• Schoolwork help: 54% used AI chatbots for schoolwork.
• Information seeking: 57% used chatbots to search for information; 64% report using chatbots overall; 47% for fun/entertainment.
• How much schoolwork: 10% do all or most with chatbot help; 21% some; 23% a little; 45% do not use chatbots this way.
• Tasks: researching a topic 48%; solving a math problem 43%; editing something they wrote 35%.
• Helpfulness: 26% say chatbots are extremely/very helpful for schoolwork; 25% somewhat; 3% little/none (among all teens; many non-users).
• Cheating perception: 59% say AI cheating happens at least somewhat often at their school (34% extremely/very often). Among schoolwork chatbot users: 76%.
• 2023→2024: ChatGPT-for-schoolwork doubled from 13% to 26%.

Who uses chatbots more (Pew demographics)

From Pew’s demographic differences chapter:

• Race/ethnicity (schoolwork help): about six-in-ten Black (~60%) or Hispanic (~62%) teens vs. roughly half of White teens (~49%).
• Intensity: Black and Hispanic teens are more likely than White teens to say chatbots are extremely/very helpful and to do all or most schoolwork with chatbot help (roughly Black 22% / Hispanic 16% / White 6% for “all or most”).
• Income: teens in households under $75K are more likely to do all or most schoolwork with chatbots (~15–20%) than those in $75K+ households (7%).
• Gender: boys and girls use chatbots at similar rates; boys are more optimistic about AI’s personal impact (41% vs. 30% positive).
• Parents vs teens: 51% of parents say their teen uses chatbots vs. 64% of teens who say they do (13-point gap). About three-in-ten parents are unsure. Just over half of parents have ever talked with their teen about chatbots; roughly four-in-ten have not. About six-in-ten parents would be OK with schoolwork use; only 18% are comfortable with emotional support/advice from a chatbot.

Common Sense Media (strong secondary)

Two related Common Sense streams—keep ages and questions separate:

• AI Use by Tweens and Teens (2026 Census) (ages 9–17): 86% use or interact with AI; 24% daily. Among AI users, 85% use it for schoolwork/homework help; 48% at least weekly for schoolwork (21% daily). Teens use schoolwork AI more often than preteens (e.g. 30% of 16–17s daily vs. 12% of 9–12s). About 73% say school/teachers communicated what they should/shouldn’t use AI for in schoolwork; fewer hear about safety (56%) or judging AI accuracy (51%).
• Teens in the AI Era: Schoolwork and Skills That Matter: 70% of teens use AI for schoolwork. Among AI users, 63% get an answer from AI (as-is 25% / rewrite to sound like them 31% / improve with what they know 35%). Only 27% say a teacher ever explained what AI is and how it works. Among schoolwork users, 44% hit a school AI block—and 59% of those switched to a personal device. 66% who use AI say it helps them understand schoolwork; 38% say it leads to fewer of their own ideas; 39% feel they’re missing learning when they use AI to finish assignments.

College Board: high school GenAI (stakeholder / policy)

From College Board’s October 2025 research release. Treat as high-school GenAI adoption—not a Pew teen replacement:

• GenAI for schoolwork rose from 79% (January 2025) to 84% (May 2025).
• Half use AI to brainstorm, edit/revise essays, and/or research sources; 69% used ChatGPT for assignments (May 2025).
• Students split on risk/benefit (~50% neither; ~25% agree benefits outweigh risks; ~25% disagree). Nearly 6 in 10 parents prefer students use GenAI than not; more than 85% of administrators see learning AI tools as valuable.
• Policy gaps: 55% of principals report no GenAI network block; ~2 in 5 schools/districts disallow student GenAI; ~1 in 5 allow use with no policy; 13% encourage in all classes; 16% leave policy to teachers; 12% to departments/grades.

Synthesis note (not a primary rate)

The Stanford HAI 2026 AI Index — Education summarizes that about four in five U.S. high school and college students use AI for schoolwork, and that only about half of middle/high schools have AI policies (6% of teachers say policies are clear). Use as context; cite Pew or Common Sense when you need a teen-specific percentage.

Sources & methods

Technical details for journalists and fact-checkers—kept out of the findings above.

• Pew (primary): Ipsos KnowledgePanel; Sept. 25–Oct. 9, 2025; 1,458 U.S. teen–parent pairs ages 13–17; margin of error ±3.3 points. Methodology · Report.
• Pew 2024 ChatGPT trend: earlier short read; n≈1,391 teens (Pew Research Center).
• Common Sense Census: SSRS Opinion Panel; Mar. 18–26, 2026; 1,204 youth ages 9–17; ±3.8 points. PDF report.
• Common Sense schoolwork brief: Youth AI Safety Institute “Teens in the AI Era” (Aug. 2026 release; nationally representative teens). Press summary.
• College Board HS GenAI: student, parent, and educator surveys June 2024–June 2025; Oct. 2025 release. High-school frame, not Pew ages 13–17.
• Stanford AI Index: education-chapter synthesis across studies—not a primary teen survey.

From K–12 habits to college AI degrees

Frequently asked questions