Most students now use AI for school, but a large share keep quiet about it. In a global survey of 2,499 students, more than half said they sometimes or often avoided revealing when they had used AI tools in their work.
The figures come from Trust, attitudes and use of artificial intelligence: A global study 2025, led by the University of Melbourne in collaboration with KPMG. Researchers surveyed more than 48,000 adults in 47 countries, including the United States, between November 2024 and mid-January 2025. A dedicated section looked at respondents who were currently studying: 65% were in university, 16% in vocational or technical programs and 18% in secondary education.
What students reported
AI use is close to universal in this group. 83% of students said they use AI in their studies on a regular basis, and about half use it weekly or daily. Almost all of them (89%) rely on general-purpose tools such as ChatGPT, Copilot or Claude, far more than tools provided by their school.
Transparency is where things get murky. Among students who use AI for school:
- 56% said they sometimes or often avoided revealing when they had used AI in their work
- 45% said they had sometimes or often presented AI-generated content as their own
- 54% had sometimes or often used AI in a course without knowing whether it was allowed
- 59% said they had used AI in ways that went against their school’s policies or guidance at least once
Part of the confusion seems to come from schools themselves. Fewer than a third of students said their education provider had a policy guiding generative AI use, about one in five said AI was banned, and a quarter did not know whether any policy existed.
The quieter cost: less learning
The survey also points to habits that can undercut the reason you are in class. 81% of students said they had put less effort into a study task at least once because they knew AI could help, and 76% had relied on AI output without checking it. About two in three had made mistakes in their work because of AI. Only about half (52%) said they regularly evaluated or verified what the tool produced.
Pressure plays a role too: half of the students said they worried about falling behind if they did not use AI. At the same time, students did report real benefits. 69% said AI had made their work more efficient, and more than half said it had reduced their workload and stress.
Keep the findings in perspective
This is a self-reported, international snapshot, not a study of American college students alone. Students were a subset of a broader adult sample, so the percentages describe this group rather than every campus. The survey also shows associations, not proof that AI use lowers anyone’s grades or skills.
How to use AI openly in your classes
None of this means you should avoid AI. It does suggest that the safest habit is to make your use visible and deliberate:
- Check each syllabus. Rules can differ from one course to the next, even within the same department. If the syllabus is silent, ask your instructor before you start, not after you submit.
- Say what you used it for. The MLA’s guidance on citing generative AI recommends citing a tool whenever you quote, paraphrase or incorporate content it created, and acknowledging functional uses such as editing or translating.
- Verify before you rely. Treat AI output like a classmate’s notes: useful for a start, but worth checking against your readings and sources.
- Do the core skill yourself. If an assignment is meant to build a skill, such as writing an argument or solving a type of problem, use AI to quiz or explain rather than to produce the answer.
If you are unsure whether something you already turned in crossed a line, your instructor or your school’s academic integrity office is the right place to ask. A quick, honest question usually costs far less than a misunderstanding later.

