Students Using AI Statistics (2026): 40+ Survey Findings

95% of full-time UK undergraduates now use AI in some form, and 94% use generative AI to help with assessed work, according to the HEPI and Kortext Student Generative AI Survey 2026. US surveys land lower but in the same direction: 85% of college students and 54% of teens report using AI for schoolwork. What students do with it is mostly explaining and summarizing, not pasting in finished essays.

Each number below lists its survey year, sample and country. A UK undergraduate panel and a US teen panel measure different things, so treat each figure as its own measurement.

How many students use AI?

Most students in the UK, the US and Latin America now use AI for school, with headline rates between 54% and 95% depending on age, country and exactly what the survey asked.

The spread in the table below is mostly about question wording. "Used AI in at least one way this year" will always score higher than "used AI as part of your learning" or "use AI daily". Read each row as its own measurement.

UK university students

• The share of UK undergraduates using AI in at least one way rose from 66% in 2024 to 92% in 2025 and 95% in 2026 (HEPI and Kortext, 2026, polled by Savanta, margin of error about 3%).
• Use of AI for assessed work went from just over half in 2024 to 89% in 2025 and 94% in 2026 (HEPI and Kortext, 2026).
• Using text generators such as ChatGPT actually fell, from 64% of students in 2025 to 56% in 2026. HEPI says the reason is unclear and suggests students may be moving to more specialized tools.
• In contrast, only 34% of UK higher education students said they used AI tools "as part of their learning" in Jisc's 2024/25 survey of 15,398 students (Jisc, 2025). The gap with HEPI's 95% is a question wording and sampling difference, not a contradiction; HEPI counts any AI use for any purpose.

US college students

• 85% of US two- and four-year college students used generative AI for coursework in the past year (Inside Higher Ed and Generation Lab, 1,047 students at 166 institutions, July 2025).
• 61% of US undergraduates use AI at least weekly, up from 42% a year earlier (Tyton Partners Listening to Learners 2026, 1,800+ undergraduates, May 2026).
• 32% of US college students use generative AI daily, less than administrators (43%) but more than faculty (25%) (Tyton Partners Time for Class 2026, spring 2026).
• Only 7% of US college students describe themselves as "all in" on AI. "Worried about dependence" was the most common stance, and 40% said they worry about depending on AI tools (Inside Higher Ed and Generation Lab, 1,038 students, May 2026).

US teens and K-12 students

• 54% of US teens aged 13 to 17 have used chatbots for help with schoolwork: 10% for all or most of it, 21% for some and 23% for a little. 45% have not (Pew Research Center, 2026, 1,458 teens, September to October 2025).
• Teens report more chatbot use than their parents realize: 64% of teens say they use chatbots, while about half of parents say their teen does (Pew Research Center, 2026).
• Use of ChatGPT specifically for schoolwork doubled among US teens in a year, from 13% in 2023 to 26% in 2024. 11th and 12th graders (31%) were more likely to use it than 7th and 8th graders (20%) (Pew Research Center, 2025, 1,391 teens, fall 2024).
• 84% of US high school students used generative AI for schoolwork in May 2025, up from 79% in January, and 69% used ChatGPT for assignments and homework (College Board, about 600 students per wave).
• AI use for homework among US students rose from 48% to 62% between May and December 2025, driven by middle and high schoolers; use among college students stayed relatively steady (RAND, 2026).

Global and Latin America

• 86% of university students used AI in the Digital Education Council's 2024 global survey of 3,839 students in 16 countries (Digital Education Council, 2024).
• 92% of students at 29 Latin American universities are actively engaging with AI (Digital Education Council, 2026, 22,941 students), and 67% use it daily or weekly (Digital Education Council, 2026).
• Only 30% of those Latin American students say their institution's current use of AI meets their expectations (Digital Education Council, 2026).

What students use AI for

Students mainly use AI to understand material: explaining concepts, summarizing readings and brainstorming come first in every major survey, ahead of generating text for submission.

• When preparing assessed work, 61% of UK undergraduates have used AI to explain concepts, 49% to summarize a relevant article, 40% to suggest research ideas and 39% to structure their thoughts (HEPI and Kortext, 2026).
• 25% of UK undergraduates have generated AI text and then edited it before submitting, and 12% have included AI-generated text directly, up from 8% in 2025 and 3% in 2024 (HEPI and Kortext, 2026).
• Across all purposes, the most common AI uses among UK students were generating text (56%), summarizing, note-taking or quizzing textbooks (38%) and editing writing with tools like Grammarly (37%) (HEPI and Kortext, 2026).
• The top three coursework uses among US college students are brainstorming ideas (55%), asking questions like a tutor (50%) and studying for exams or quizzes (46%). Fewer use it to complete assignments (25%) or write full essays (19%) (Inside Higher Ed and Generation Lab, July 2025).
• Among US undergraduates, 37% use AI for help with written assignments, 30% for tutoring and 30% to summarize lectures (Tyton Partners Listening to Learners 2026).
• US students aged 12 to 29 most often use AI to get better explanations of assignments (38%), brainstorm (35%), look up facts (33%) and draft or revise writing (33%). ChatGPT was the most used tool (53%), and Google Gemini use (28%) more than doubled between May and December 2025 (RAND, 2026).
• About four in ten US teens have used chatbots to research a topic or solve math problems, and 35% to edit something they wrote (Pew Research Center, 2026).

AI for studying vs AI for writing

Students themselves draw the line at AI doing the work: they widely accept AI for explanations and research, and mostly reject it for writing essays or getting answers.

• 54% of US teens say it is acceptable to use ChatGPT to research new topics, 29% say the same for solving math problems and only 18% for writing essays. 42% say essay writing with ChatGPT is not acceptable (Pew Research Center, fall 2024).
• Getting direct answers to homework was the one AI use that many US students saw as cheating: 45% called it cheating. Nearly 80% said using AI to understand an assignment was not cheating, as did 72% for brainstorming and 67% for looking up facts (RAND, December 2025).
• 65% of Latin American university students worry AI may make learning too shallow and discourage critical thinking and creativity (Digital Education Council, 2026).
• Concern about critical thinking is higher among students who avoid AI: 78% of non-users agreed AI harms critical thinking, compared with 60% of users. Female students were more concerned than male students (75% vs 59%) (RAND, 2026).
• 49% of UK undergraduates say AI has made their student experience better, while 16% say it has made it worse and 35% say it has had no significant impact (HEPI and Kortext, 2026).

The pattern matches what learning research would predict. Asking AI to explain a concept and then testing yourself on it keeps the effortful part of learning with the student. Asking it to write the answer removes it. Our guide to the best ChatGPT prompts for making flashcards covers how to use a chatbot for the first kind of task, and our guide to AI study tools for students compares tools built around it. If you want a tool that makes you do the recall rather than skip it, StudyCards AI turns lecture PDFs into flashcards you still have to answer yourself.

AI, cheating and academic integrity

There are three different numbers people call "AI cheating": what students admit to, what students believe others do, and what institutions actually catch. They differ by an order of magnitude, so it matters which one you cite.

Self-reported

• 12% of UK undergraduates admit including AI-generated text directly in assessed work (HEPI and Kortext, 2026). This is not necessarily misconduct, since some courses now permit or require AI use.
• 19% of US college students say they have used AI to write full essays and 25% to complete assignments (Inside Higher Ed and Generation Lab, July 2025).
• For years before ChatGPT, 60% to 70% of US high school students reported at least one cheating behavior in the previous month. In Stanford researchers' 2023 surveys, which added AI-specific questions, that share stayed about the same or dipped slightly (Stanford Graduate School of Education, 2023).

Perceived

• 59% of US teens think using AI to cheat happens at least somewhat often at their school, including about a third who say extremely or very often. 14% say it rarely or never happens and 15% are unsure (Pew Research Center, fall 2025).
• Among teens who have used chatbots for schoolwork, 76% say students at their school use chatbots to cheat at least sometimes (Pew Research Center, 2026).
• 61% of Latin American university students fear AI misuse by their peers (Digital Education Council, 2026).
• US college students blame pressure to get good grades (37%) as the top driver of academic integrity violations, ahead of time pressure (27%) and not caring about integrity policies (26%) (Inside Higher Ed and Generation Lab, July 2025).

Detected and proven

• UK universities recorded almost 7,000 proven cases of cheating with AI tools in 2023-24, or 5.1 per 1,000 students, up from 1.6 per 1,000 in 2022-23. Figures to May 2025 pointed to about 7.5 per 1,000 for 2024-25 (The Guardian, 2025, Freedom of Information requests to 155 universities, 131 responded).
• More than 27% of responding UK universities did not yet record AI misuse as a separate misconduct category in 2023-24, so the proven figure is an undercount (The Guardian, 2025).
• In its first year, Turnitin's AI detector reviewed over 200 million papers. About 11% had at least 20% AI writing present and about 3% had at least 80% (Turnitin, April 2024). These are detector flags, not findings of misconduct.

Can AI use be detected?

• Researchers at the University of Reading submitted fully AI-written answers to real undergraduate psychology exams. 94% went undetected, the AI answers averaged half a grade boundary higher than real students', and there was an 83.4% chance the AI submissions on a module would outperform a matched set of real students (Scarfe et al., PLOS ONE, 2024).
• Seven popular GPT detectors misclassified human-written TOEFL essays by non-native English speakers as AI-generated at an average rate of 61.22%. 97.8% of the 91 essays were flagged by at least one detector, while essays by US 8th graders were classified almost perfectly (Liang et al., Patterns, 2023).
• 42% of UK undergraduates say fear of being accused of cheating makes them less likely to use AI, down from 53% in 2025. It is still the most common deterrent, ahead of hallucinations (35%) (HEPI and Kortext, 2026).

Teacher and institution policies

Policy is lagging use: most students say AI rules are set teacher by teacher rather than by the institution, and many institutions still have no formal policy.

• Only about one-third of US students said their school had a schoolwide AI policy; many, especially college students, said the rules varied by teacher (RAND, December 2025).
• Just 32% of US college administrators report an institution-wide AI policy. Where one exists, only 22% of faculty think it is effective, compared with 44% of administrators (Tyton Partners Time for Class 2026).
• Only 13% of faculty at US colleges that ban AI entirely say their policy is effective, compared with 30% where AI is integrated into assignments (Tyton Partners Time for Class 2026).
• In a May 2023 UNESCO survey of just over 450 schools and universities worldwide, fewer than 10% had institutional policies or formal guidance on generative AI (13% of universities, 7% of schools), and only two reported a full or near-full ban (UNESCO, 2023).
• 39% of US high school and district administrators said their school or district does not allow students to use generative AI, 19% allow it with no policy in place, 16% leave it to individual teachers and 13% encourage it in all classes (College Board, 2025, about 1,300 administrators, June 2025).
• 65% of UK undergraduates say the way they are assessed has changed a lot in response to generative AI, up from 56% in 2025 (HEPI and Kortext, 2026).
• The share of UK universities providing AI tools to students rose from 9% in 2024 to 23% in 2025 and 38% in 2026, while 50% of students think their institution should provide them (HEPI and Kortext, 2026).
• 68% of UK undergraduates say AI skills are essential to thrive today, but only 48% say teaching staff help them develop those skills (HEPI and Kortext, 2026).
• 97% of US college students want their institution to respond to academic integrity threats from AI, but only 21% favor AI-detection software. 53% want education on ethical AI use instead (Inside Higher Ed and Generation Lab, July 2025).
• A quarter of US public K-12 teachers (25%) said AI tools do more harm than good in K-12 education, and only 6% said more good than harm. High school teachers were the most negative, at 35% (Pew Research Center, 2024, 2,531 teachers, fall 2023).
• 56% of Latin American university students are concerned about data privacy when using AI tools (Digital Education Council, 2026). Our explainer on AI data privacy for students covers what happens to notes you upload.

Methodology

We included only primary sources: official survey reports and publisher pages (HEPI, Pew Research Center, RAND, College Board, Tyton Partners, Digital Education Council, Jisc, UNESCO, Inside Higher Ed's own Student Voice survey), peer-reviewed studies, and the Guardian's own Freedom of Information dataset. Every figure was checked against the source in September 2026, and we used the latest edition of each recurring survey. Stats aggregator sites were excluded. Because samples, ages, countries and question wording differ, we report each number with its survey and do not average or rank across surveys. For the broader picture on AI study tools, see our complete student guide to AI flashcards.

Sources

1. Stephenson, R. and Armstrong, C. (2026). Student Generative AI Survey 2026 (HEPI Report 199). Higher Education Policy Institute and Kortext.
2. McClain, C., Anderson, M., Sidoti, O. and Bishop, W. (2026). How Teens Use and View AI. Pew Research Center.
3. Sidoti, O., Park, E. and Gottfried, J. (2025). About a quarter of U.S. teens have used ChatGPT for schoolwork, double the share in 2023. Pew Research Center.
4. Lin, L. (2024). A quarter of U.S. teachers say AI tools do more harm than good in K-12 education. Pew Research Center.
5. Schwartz, H. L. and Diliberti, M. K. (2026). More Students Use AI for Homework, and More Believe It Harms Critical Thinking: Selected Findings from the American Youth Panel (RR-A4742-1). RAND Corporation.
6. RAND Corporation (2026). Student Use of AI for Homework Rises as Concerns Grow About Critical Thinking Skills. Press release, March 17, 2026.
7. College Board (2025). U.S. High School Students' Use of Generative Artificial Intelligence. Research brief.
8. College Board (2025). New Research: Majority of High School Students Use Generative AI for Schoolwork. College Board Newsroom.
9. Flaherty, C. (2025). Survey: College Students' Views on AI. Inside Higher Ed Student Voice survey with Generation Lab.
10. Flaherty, C. (2026). Why Students Aren't All In on AI, and What They Want From Colleges. Inside Higher Ed Student Voice survey with Generation Lab.
11. Tyton Partners (2026). Time for Class 2026: The AI Tipping Point: From Monitoring Students to Engaging Them.
12. Tyton Partners (2026). Tyton Partners Releases Listening to Learners 2026. Press release, September 15, 2026.
13. Digital Education Council (2026). 92% of Students and 79% of Faculty Actively Engaging with AI: Findings from AI in Higher Education LATAM Survey 2026.
14. Digital Education Council (2026). AI Adoption Is Nearly Universal Among Students, But Confidence Is Not.
15. Digital Education Council (2024). Digital Education Council Global AI Student Survey 2024.
16. Jisc (2025). 2024/25 UK Higher Education Students Digital Experience Insights Survey Findings.
17. UNESCO (2023). UNESCO survey: Less than 10% of schools and universities have formal guidance on AI.
18. Goodier, M. (2025). Thousands of UK university students caught cheating using AI. The Guardian.
19. Scarfe, P., Watcham, K., Clarke, A. and Roesch, E. (2024). A real-world test of artificial intelligence infiltration of a university examinations system: A "Turing Test" case study. PLOS ONE, 19(6), e0305354.
20. Liang, W., Yuksekgonul, M., Mao, Y., Wu, E. and Zou, J. (2023). GPT detectors are biased against non-native English writers. Patterns, 4(7), 100779.
21. Turnitin (2024). Turnitin marks one year anniversary of its AI writing detector with millions of papers reviewed globally. Press release, April 9, 2024.
22. Spector, C. (2023). What do AI chatbots really mean for students and cheating? Interview with Victor Lee and Denise Pope. Stanford Graduate School of Education.

Cite this page

You are welcome to use these statistics in your own writing. Please credit StudyCards AI with a link to this page.

APA

Groves, C. (2026). Students Using AI Statistics (2026): 40+ Survey Findings. StudyCards AI. https://studycardsai.com/blog/students-using-ai-statistics

HTML link

<a href="https://studycardsai.com/blog/students-using-ai-statistics">Students Using AI Statistics (2026): 40+ Survey Findings</a> Copy HTML

Frequently Asked Questions

What percentage of students use AI?

It depends on the survey and the age group. 95% of full-time UK undergraduates used AI in at least one way in HEPI and Kortext's 2026 survey (1,054 students, December 2025). In the US, 85% of college students used generative AI for coursework in the past year (Inside Higher Ed, July 2025), and 54% of teens aged 13 to 17 have used chatbots for schoolwork (Pew Research Center, fall 2025).

How many students use ChatGPT for schoolwork?

Among US teens aged 13 to 17, 26% said they had used ChatGPT for schoolwork in fall 2024, double the 13% in 2023 (Pew Research Center, 1,391 teens). Among US high school students specifically, 69% reported using ChatGPT for assignments and homework in May 2025 (College Board). ChatGPT is the most used tool in almost every survey.

What percentage of students use AI to cheat?

No survey measures cheating directly. In HEPI's 2026 UK survey, 12% of undergraduates admitted pasting AI-generated text directly into assessed work. Proven AI misconduct cases at UK universities were far lower, 5.1 per 1,000 students in 2023-24 (Guardian FOI data). In Pew's fall 2025 survey, 59% of US teens believed AI cheating happens at least somewhat often at their school.

Can teachers and AI detectors tell if students used AI?

Often not. In a University of Reading test published in 2024, 94% of fully AI-written exam answers went undetected by markers. Detectors also misfire: a Stanford-led study found seven GPT detectors wrongly flagged 61.22% of 91 human-written TOEFL essays by non-native English speakers as AI-generated (Liang et al., 2023).

Do schools and universities have AI policies?

Mostly not at the institution level. Only about one-third of US students said their school had a schoolwide AI policy (RAND, December 2025, 1,214 students), and just 32% of US college administrators reported an institution-wide AI policy (Tyton Partners, spring 2026). In a 2023 UNESCO survey of about 450 schools and universities, fewer than 10% had formal guidance.

What do students use AI for most?

Explaining things, more than writing. In HEPI's 2026 UK survey, 61% of undergraduates had used AI to explain concepts for assessed work and 49% to summarize an article. In Inside Higher Ed's July 2025 US survey, the top uses were brainstorming (55%), asking questions like a tutor (50%) and studying for exams or quizzes (46%).