Last fall, at 11:42 pm, a student emailed me in a panic asking for an extension on his paper. I gave him another day. The next afternoon, when I was cleaning up my office, he came in smiling and handed me the finished essay. “I was busy last night,” she said. “I almost wrote Chat. I’m glad I didn’t.”
He would not even share a moment of temptation—and relief. It helped me see clearly that, in the age of AI, we need to prevent cheating not by increasing student monitoring, but by increasing student support.
Many colleges have done the opposite. Feeling under siege by AI-assisted cheating, they have tightened the perimeter: strict policies, zero-tolerance language, AI detectors, locked browsers, keystroke tracking, verbal defense, blue-book tests.
The answer is understandable. Research has long suggested that academic misconduct is not uncommon: More than 60 percent of university students admit to cheating in some way, and nearly one in five undergraduates admit to cheating repeatedly. And now, with artificial intelligence, cheating is not only easier but harder to spot.
So, should colleges double down on admissions? Even if AI detectors were reliable, they would no longer be effective, because they broadcast a message to students: We think you’re cheating.
This oversight makes academic relationships worse. It sets the class up like a game of cat and mouse: You try to cheat, we try to catch you. Actually doing the work, at that time, becomes a losing position, when the student simply failed to find an effective shortcut.
In the classroom, that suspected harmony creates anxiety. A recent YouGov survey conducted by Studiosity and reported on Within Higher Ed found that 75 percent of students using AI reported stress about being flagged incorrectly for cheating.
If student welfare is one of the primary concerns in higher education, as it should be, then colleges should be concerned not only with curbing bad behavior, but also with policies that reinforce anxiety and division. Research on student persistence suggests that emotional stress and mental health difficulties are the biggest reasons students consider leaving college. Some studies link student burnout and disengagement to academic performance and greater intention to drop out. And research suggests that when students aren’t sure they belong in an academic environment, common obstacles can be evidence, in their minds, that they don’t belong in college at all.
An approach that focuses on support rather than supervision sends a message of reassurance. Say to students: Yes, shortcuts exist, but they are bad deals. We trust you, we want you to learn and we have designed this course to help you succeed.
In my classes, I don’t think today’s students want cheaters. In casual conversations with them, I found that they really care about their education. This is in line with the YouGov/Studiosity survey finding that only 21 per cent of students said they would rely entirely on AI to write their papers even if it was allowed. Many students are not looking to give their thinking to the machine.
In order to deal with student cheating, instead of trying to discourage it with threats and accusations, we must weaken the conditions that tempt students to cheat in the first place. Psychology helps identify disciplines that increase or decrease cheating.
Another factor that causes cheating is a lot of ignorance: the phenomenon where students who value honesty are more tempted to cheat when they suspect that cheating has become the norm, lest they follow their peers.
So a course that emphasizes supervision not only fails to prevent cheating; it can be really inspiring. If pedagogy leans toward discovery over development, it cultivates a class ideology of blame rather than community.
Course redesign, in contrast, works against many ignorances by demonstrating the opposite: that most students are expected to participate honestly, and that the conditions for doing so are actually in place.
Research suggests that cheating is often less about the student than about the student’s circumstances. Scholars such as Eric Anderman and David Rettinger have shown that academic misconduct is largely influenced by context: pressure, competition, peer norms, student motivation and whether the course emphasizes grades over learning. Students are not always “imposters” as a stable identity; rather, many are honest students who make bad decisions under stress, pressure or poorly designed incentives.
Behavioral research helps explain why those weak moments arise. The temptation to cheat usually depends on three factors: motivation, measurement and opportunity. Remove or weaken any one of them and the stool begins to row. Course redesign can do just that.
Motivations to cheat, the first condition, are often driven by depression—as in my student’s case. When it’s late at night and a student is faced with a choice between submitting an AI-generated paper or getting a zero, cheating may seem like the logical choice. As one professor later put it Within Higher Ed“It’s easy to blame the students, but when it’s 9:45 at night and you have an assignment due in 15 minutes and you just finished your shift at work and you’re tired, it’s very easy and very tempting to take that question to AI.”
One solution is to create pressure valves: two no-questions-asked extension tickets, one major redo per semester or a grace window after the official deadline when students can revise and resubmit. Students rarely cheat when they see ways to recover, and they are happy when teachers care enough to give them room to breathe.
Rationalization, the second criterion, is the story students tell themselves to justify cheating. If students see assignments as busy work—disconnected from their lives, goals, or any important skill—they will be more likely to outsource it to AI. Why not cheat when work feels like another pointless hoop?
To prevent that idea, the faculty must answer three questions beforehand: What is the ability of this structure? How does it work without this course? Why does it matter?
Opportunity, the third criterion, can be reduced by shifting teaching from product to process. When the grade of a course depends on a few weighty papers, a simple AI button strikes like an open invitation. But when messy academic work is highlighted and valued—through annotated frameworks, review histories, conferences, drafts and written reflections on change—opportunity spreads, ownership diminishes and students become more invested in their work.
This shift in process also eases the pressure on students who are already struggling: those with ADHD who struggle with time management, those in the age group who may need more structure than open-ended information provides and multilingual students who need repeated feedback rather than one high-stakes test. Adding a large assignment to a proposal, draft and revision—where each part counts less—spreads the pressure throughout the semester, leaves room for feedback and cultivates a sense of pride in the writer.
This is also where AI can be registered as an alliance. A course-specific chatbot, introduced in advance and trained with course materials, can provide 24-7 support. Students can ask questions without fear of judgment, clarify directions, gain a better understanding of the material and get a first-rate answer. Encouraging students to use AI in this way reinforces the lesson’s implicit message: Support is available, shortcuts are not needed and the point is your progress.
None of this eliminates cheating. Some students will cheat no matter what you do. But tightening the course to catch cheaters may catch a few, while rebuilding that course to support students will reach the majority: those who want to learn with integrity and purpose.
Academic integrity will not be protected by better detectors, but by better design. We need to replace the tools of inquiry with supportive conditions that encourage honest effort. A little extension gave my student some breathing room to complete the work that felt right. That combination is rarer than it should be, and more powerful than any detector.
