AI Proofing Your Classroom (Sort Of)

And other

Okay, I’ll come right out and say it: this title is a bit deceptive. “AI-proofing” sounds sleek and trendy, the kind of phrase that might catch your eye on a conference schedule or podcast episode. But what I’m actually advocating for is far less flashy and far more important: helping students build the critical thinking and communication skills they’ve always needed—long before ChatGPT entered the scene.

In this post, I want to share a few ways I’ve been reworking my online classes in light of generative AI. These changes aren’t about playing whack-a-mole with cheating or making my assignments “AI-resistant.” Instead, I’m re-centering my pedagogy around learning rather than credentialing—something scholars like

John Warner

and

Jesse Stommel

have long argued for. In short, I’m asking: what would happen if we stopped designing assignments with suspicion in mind and started designing for student agency and growth?

Understanding Why Students Misuse AI

One of the most common concerns I hear from educators is: “My students are using AI to cheat; how do I stop it?” But that’s not the most productive question we can be asking. Instead, we should be asking: Why do students feel the need to outsource their work to AI in the first place?

A 2009 study published in the Journal of Business Ethics explored this very question. Researchers surveyed students on their motivations for cheating and found that over 60% of both business and non-business majors admitted to it. The top reason? A “desire to get ahead.” Not laziness. Not hatred of the subject matter. A sense of urgency and competition in a system that often values grades over growth.

This insight changed how I think about AI use in my classroom. If students are primarily motivated by fear of falling behind or pressure to achieve, then cracking down with harsher rules won’t solve the root problem. Instead, educators can move towards creating a learning environment where students don’t feel the need to cut corners because they’re engaged, supported, and genuinely learning.

1. Rethinking “Busy Work”: Speaking Instead of Writing

Let’s talk about discussion boards.

I’ll be honest: I’ve always had mixed feelings about them. But after the rise of AI tools, my skepticism deepened. If students can ask ChatGPT to summarize a reading and generate a post in under a minute, what are we even measuring? Worse, if the format itself doesn’t feel meaningful, students are more likely to disengage—whether they’re using AI or not.

This summer, I replaced weekly discussion posts with short “video check-ins.” Students record 5-minute reflections on their readings, share key takeaways, raise questions, and connect concepts to their lives. These videos are posted to a class Canvas discussion board where peers can comment or respond.

The impact has been huge, and I get a better sense of their authentic voices and thinking. It’s harder to fake comprehension in a video than in a text box, and as a bonus, students are building confidence with basic digital communication tools like creating a video for their peers—a critic

al skill in today’s media landscape.

2. Co-Creating AI Policies With Students

One of the most transformative shifts I’ve made is moving away from a top-down approach to academic integrity and instead co-creating classroom AI policies with students.

This summer, I invited my students to define what ethical and unethical use of AI looks like in our course. Using Padlet and small group discussions, they shared perspectives on when AI tools feel helpful (e.g., brainstorming, grammar support) and when they cross a line (e.g., generating full essays or citations without comprehension). Together, we developed a shared agreement that emphasized transparency, critical thinking, and learning over performance.

The result? Buy-in. When students have a voice in setting expectations, I believe they’re more likely to honor them. More importantly, these conversations modeled the very kind of digital literacy I hope they take into other classes and life beyond school.

(If you want the full breakdown of this model, I have a dedicated Substack post on AI policy co-creation).

3. Revisiting Flipped Classrooms

After a recent post from

Nick Potkalitsky

, I’ve also found myself revisiting the flipped classroom model—not as a “new” idea, but as a framework that aligns beautifully with the presence of AI.

In a flipped classroom, students encounter foundational content (readings, videos, lectures) before class time. Then, synchronous sessions are used for deeper analysis, peer work, or applied practice. Essentially, anything that benefits from live interaction and can’t be easily outsourced to a chatbot.

When it comes to AI, flipped classrooms offer two advantages. First, they front-load low-stakes learning, which students can supplement with AI tools in transparent ways (e.g., asking AI to explain a difficult concept or generate an outline). Second, they preserve class time for the kind of work AI can’t do well: authentic discussion, debate, collaboration, and reflection.

This model positions AI as a potential partner in the learning process, not a threat to be neutralized. It also allows me to assess learning in real-time, with assignments that privilege voice, originality, and connection.

Final Thoughts

None of this is really about “AI-proofing” in the literal sense. If a student really wants to use ChatGPT to write a paper behind my back, they’ll probably find a way to do it. But by shifting my classroom culture—from compliance to curiosity, from busy work to authentic expression—I’ve found that fewer students even want to go that route.

AI is pushing us to reconsider how we teach, but maybe that’s not a bad thing. Maybe it’s the push we’ve needed to finally let go of outdated practices, ask better questions, and build classrooms that are less about getting ahead and more about growing together.

If you’ve tried any of these approaches—or are considering them—I’d love to hear from you in the comments. How are you rethinking your classroom design in the age of AI?