Cheating Isn’t New. But AI Has Changed the Game
by Vicki
Let’s start with an uncomfortable truth: cheating is not new. It’s practically a time-honored tradition in academia. Whether it was copying a friend’s homework, writing answers on a wrist, or slipping crib notes into a sleeve—every generation has found its shortcuts.
But now, the shortcut is absolutely amazing comparatively speaking.
Today’s AI tools can generate essays, solve equations, and even mimic a student’s voice in writing. And understandably, that alarms a lot of educators, especially those of us who built our careers on critical thinking, originality, and the slow, painful processes involved in learning.
I get it. I was a professor for decades. I still wince when students outsource their own growth. But I also know this: the tools change. But the challenge remained the same.
The Old Problem, Upgraded
When I first started teaching at the university level, creating exams meant making our own copies, literally. If I wasn’t doing it myself, then a secretary or graduate student was. We’d head to a shared copy room with a stack of originals, wait our turn at the machine, and do our best to keep track of every page.
But let’s be honest, there wasn’t much security. Paper exams were often left unattended, and misprints or jammed copies regularly ended up in the trash. I heard one story, from a former student about a janitor who used to selling discarded exams.
And that was just one layer of the problem.
Some fraternity and sorority houses reportedly had filing cabinets filled with old tests and assignments, passed down like heirlooms. If you’re old like me, you’ve had to consider whether your students should even be allowed to keep their graded exams or assignments. It’s a great resource that can be used in preparation of a comprehensive final exam. Or, it’s a valuable commodity that may be sold (or shared freely) around campus, or even placed online.
Even after we moved to asynchronous online courses, the worries didn’t go away. Screenshots. Group chats. Shared answer banks. Cheating didn’t disappear. It just changed costumes.
Now, with AI, it’s not just about protecting test questions. It’s about rethinking what authentic assessment even looks like.
What’s different today isn’t the temptation, it’s the plausibility. AI makes cheating faster, easier, and harder to detect. That’s unsettling.
But maybe we’re asking the wrong questions.
Cheating Isn’t Just Real, it’s Relatable
Let’s be honest: cheating isn’t just something students do. It’s something they participate in as a part of a cultural norm. It’s part of pop culture, comedy, and coming-of-age storylines. There are entire movies built around it.
From The Perfect Score to 21, from sitcom gags to TikToks, cheating is portrayed as clever, funny, even justified, especially when the system is viewed as rigid or unfair. In high school and college movies, the test isn’t just academic, it’s social and personal.
That matters. Because it means we’re not just dealing with student behavior, we’re dealing with a student narrative. One that says, “Cheating is how you survive school when the odds are stacked against you.” In that narrative, as educators, we end up being the bad guy, by the way.
If we want to change the behavior, we may also need to change the story.
Why Do Students Cheat?
It’s easy to look at cheating and feel disappointed. But it’s more helpful to look at it with curiosity.
Research shows that students don’t usually cheat because they’re lazy or malicious. They cheat because they’re pressured, overwhelmed, or disconnected.
👉 Why are they tempted to cheat in the first place?
Our students are under pressure. What kind of pressure? Isn’t this a carefree time of life for them? Many of them are under academic, financial, and social pressures. They often feel overwhelmed. Cheating becomes a shortcut when students believe the stakes are too high to fail and the support isn’t enough to succeed.
👉 What kind of assignments are so generic, a robot can do them?
Often, it’s not the assignment itself—it’s the deliverable. Essays and discussion board posts, for example, are easy for AI to replicate. That doesn’t mean the topic lacks value, but it does mean the format might need a refresh.
👉 How can we design learning experiences that AI can’t fake—because they tap into something deeply human?
We can require students to show their process, bring in personal context, or engage in real time.
✋ What Can Only a Student Give You?
AI is excellent at producing tidy, fluent responses. But there are still plenty of deliverables that only a real, thinking student can authentically produce:
- A messy draft and revision trail
Let students show their thinking process. Have them offer notes, edits, and questions. AI can fake fluency, but not development. - A spoken explanation
A short video or live check-in reveals real understanding far more than polished paragraphs. - A personal connection
Ask students to relate a concept to their life, community, or beliefs. AI can mimic tone, but our students have experiences. We may have to tweak how they relay that information to us. - Real-time collaboration
Group activities and peer responses reveal participation that AI can’t easily impersonate. - Reflective commentary
Ask: “Did you use AI? How? What did you keep? What did you change?” Reflection reveals engagement. Let students when the assignment is provided that you anticipate the use of AI (as a tool, where possible) and want a journal record type entry of how it was used. Copies of prompts, etc. Collect and share prompts and the responses that emerged. Discuss that. - Local or specific sources
Interviews, campus events, community stories can be great resources for students and can ground assignments in real, AI-proof material. - Sketches, diagrams, or concept maps
Especially in tech and design-heavy subjects, visuals show originality AI still struggles to fake.
Rethinking the Classroom Experience
One of the best older strategies I’ve seen, that should be re-introduced is “Flipping the classroom”.
In a flipped model, students encounter the core material like videos, readings, or short lectures before class. That leaves classroom time for what matters most: discussion, application, analysis, and creative work. In short, all the things AI can’t do for them.
Flipping the classroom doesn’t just change when students engage with content, but it changes how they engage. It moves them from being passive recipients to becoming active participants. It’s harder to fake understanding when you’re applying a concept in real time, contributing to a group discussion, or critiquing a peer student’s project.
And let’s be honest: it’s harder to cheat when you’re part of something living, interactive, and grounded in the room you’re in.
It may take more preparation up front. But it’s a proven way to foster real learning and make AI a tool within the process, not a replacement for it.
But What About Large Class Sizes?
You may be thinking, “This sounds great for small seminars or small classes but what about when I’m teaching 200 students at a time?”
It’s a fair question. And to be honest, I’ve never been a fan of really large class sizes not as a student or as faculty. The scale may be efficient for enrollment targets, but it makes real human connection and meaningful assessment harder to achieve. However, I am still convinced we have to find a way to change how we interact with our students, even in this environment.
And here’s the thing: this conversation about cheating and assignment design should also push us to rethink attendance policies and grading structures.
If attendance is based on passive presence, just showing up physically or logging in, then students may disengage while still technically meeting expectations. But when participation means something, students know that their presence matters. We want more than simply a warm body present. We want a measurable interaction. That’s one of the main ways we can thwart AI cheating.
Grading systems that reward memorization or favor high-stakes, one-shot performance over consistent engagement create the kind of stress that drives dishonest behavior.
We may not always have the power to change class size. But we can advocate for better design within those constraints: smaller discussion groups, layered assignments with checkpoints, or optional reflective add-ons. Every small move toward presence and personalization matters.
Teaching Is Not Policing
Most of us did not become educators because we wanted to be detectives. I despise interrogating students, scanning for signs of deception, and second-guessing every assignment turned in. That’s a miserable experience for everyone involved.
For me, teaching has always been about learning and discovering together. I don’t mind if AI helps me or my students explore something more thoroughly. In fact, I welcome the curiosity it can spark. I just don’t want my students to miss the deeper opportunities available. I want them to experience those moments when they surprise themselves. I love the “light bulb coming on” moments that I have experienced in my life. I love being present when those moments happen for my students. I want them to experience those moments, even when we’re not together. I don’t mind if they use AI for that purpose. I want them to learn. I want my students to love to learn.
We can’t eliminate cheating entirely. But we can create environments where cheating feels irrelevant or unnecessary because the work feels meaningful, the expectations feel fair, and the classroom feels like a place worth showing up for.
Standing Square with Students
If you feel a little out-of-step with the latest tech, but you still want to lead with clarity and compassion, let me know how I can help you with that pursuit.
Most of us became educators because we love our discipline and we love learning. And many of us are still, at heart, students ourselves. That means we’re not just reacting to AI, we’re also learning alongside it. I don’t want a “fear of the unknown” to steal that from us.
We want a healthy relationship with learning for ourselves, and for our students. That means embracing tools that help us think more deeply, communicate more clearly, and explore more bravely. AI can be part of that process. It just shouldn’t replace it.
So if you’re wrestling with how to integrate AI into your teaching without losing what matters most you’re not alone. And you’re not behind. You’re just standing at a new starting line.
Let’s figure this out—together.