What faculty are afraid of
Every fresh term, university faculty face some version of the same question: are students learning, or just moving through the motions of a degree? This year that question has a new backdrop. A November 2025 survey of 1,057 faculty by the American Association of Colleges and Universities and Elon University found that 90% believe generative AI will diminish students' critical thinking skills, 83% believe it will shorten their attention spans, and 95% expect it to increase their overreliance on the tools themselves. Nearly three in four, 73%, have personally dealt with an academic integrity issue involving AI. These aren't fringe worries from a resistant minority. They're the majority view of the profession, and they're specific: not "AI is disruptive" in the abstract, but a shared sense that something measurable is happening to how students think.
What students are already doing
Students aren't hiding from this either. HEPI's 2026 Student Generative AI Survey, based on 1,054 UK undergraduates, found that 95% now use AI in some form and 94% use it for assessed work. Only 12% admit to submitting AI-generated text directly, though that figure has quadrupled since 2024, when it stood at 3%, which suggests the boundary between "using AI to help" and "letting AI do it" is moving fast. Sitting two of the survey's own quotes side by side tells its own story. One student described using AI to summarise dense readings so they could "focus on critical analysis and deeper understanding." Another wrote, simply, "I'm not using my brain at all." Same tool, same term, two entirely different learning experiences, and the survey gives no indication that either student was doing anything unusual.
Why the mechanism hasn't changed
What separates those two students isn't the tool. It's what the classroom asked of them once they'd used it. AI can answer the "what" faster than any lecture can, faster than any textbook, faster than office hours. What it can't do is stand in for a student at the moment they're asked to predict an outcome, defend a choice, or explain something back in their own words, because those moments only work if the student is the one doing the retrieving. This is the older idea underneath all of this: retrieval practice, sometimes called the testing effect, the well-established finding that actively pulling information out of memory builds far stronger, longer-lasting understanding than simply being given the information again. It's not new, and it's not about AI. It's just that AI makes it unusually easy to skip.
A 2025 Drexel University study puts a number on that gap. Researchers gave data science students AI-generated retrieval practice questions during lecture breaks, low-stakes prompts that forced students to recall and apply what they'd just been taught, and compared their later quiz performance to a control group that received no such practice. The retrieval group scored 89%. The control group scored 73%. A 16-point difference the researchers called both statistically and practically significant, from changing nothing about the content, only whether students were made to retrieve it. That finding hasn't changed because AI exists. If anything, it explains exactly what's at stake: the mechanism that produces real learning still runs on effortful recall, and AI's usefulness for a student depends entirely on whether it's used to prompt that effort or to replace it.
Where policy runs out
This is why questioning technique, not AI policy, may turn out to be higher education's most under-used lever. In the AAC&U survey, 87% of faculty said they'd written explicit AI-use policies, yet 59% still said their institution wasn't well prepared to help students face what's coming. Policy is necessary, but it governs what a student submits after the fact. It does very little for the ninety minutes before they submit it, in the lecture hall or seminar room itself, where the actual thinking either happens or doesn't, and where no policy document has ever made anyone think harder.
None of this is an argument for resisting AI in the classroom, and the data doesn't support that conclusion either; plenty of students are already using it well. It's an argument for being far more deliberate about which moments still belong to the student alone, and designing lectures, seminars and assessments to protect those moments on purpose, the way this issue's Stories Behind the Slides shows one instructor already doing, one retrieval question at a time.
References
American Association of Colleges and Universities, & Elon University's Imagining the Digital Future Center. (2026, January 21). National survey: 95% of college faculty fear student overreliance on AI and diminished critical thinking among learners who use generative AI tools. AAC&U. https://www.aacu.org/newsroom/national-survey-95-of-college-faculty-fear-student-overreliance-on-ai-and-diminished-critical-thinking-among-learners-who-use-generative-ai-tools
An, Y., Liu, J., Acharya, N., & Hashmi, R. (2025). Enhancing student learning with LLM-generated retrieval practice questions: An empirical study in data science courses. arXiv. https://arxiv.org/abs/2507.05629
Stephenson, R., & Armstrong, C. (2026, March 12). Student generative artificial intelligence survey 2026 (HEPI Report 199). Higher Education Policy Institute. https://www.hepi.ac.uk/reports/student-generative-ai-survey-2026/
