Generative AI is increasingly used by students for schoolwork, with a Policy Institute study in the UK indicating a significant uptake. However, pinpointing the exact ways AI is being utilized and its long-term impact on learning remains unclear. Concerns about cheating have prompted some institutions to attempt bans or require disclosure of AI use, while others are developing training and guidelines to prepare students for an AI-integrated workplace, recognizing that restricting exposure could disadvantage them.
A 2023 study by RM Technology found that two-thirds (67%) of secondary school students admit to using generative AI. Their primary uses include researching (42%), summarizing (41%), checking work (24%), brainstorming essay topics (18%), and taking tests (18%). Despite these reported uses, many students express concern that AI might be hindering their critical thinking abilities.
A RAND Corporation survey of middle, high school, and college students revealed that nearly 7 in 10 are worried about AI eroding their critical thinking skills. This concern grew significantly among middle schoolers from 48% in February 2025 to 68% by December, and among high schoolers from 55% to 65% in the same period. College students also voiced apprehension, with 70% believing AI use could harm their critical thinking. Even with these worries, AI usage for homework rose from 30% to 46% for middle schoolers and from 49% to 60% for high schoolers between May and December 2025.
While students' definition of cheating is narrow, with most not considering AI use for tasks as cheating, 45% did view using AI for homework answers as unethical. High school students, in particular, are more concerned about accusations of cheating (64%) and perceive rules for AI use as dependent on individual teachers. Experts like Heather Schwartz of RAND warn against students using AI for "first-draft thinking," suggesting that it bypasses the crucial initial struggle that fosters cognitive development, potentially shortchanging the learning process.
