Cheating Isn’t New—And It’s Not the Root Problem

Before you stop reading—no, I’m not under the illusion that students aren’t cheating with AI. Students have always cheated, and they will continue to do so with whatever tools are available. In fact, research suggests that rates of cheating have not dramatically changed in the age of AI (Stanford Graduate School of Education).

Stanford researchers remind us that cheating long predates AI and encourage us to consider why students cheat. The reasons are familiar: high pressure, heavy workloads, limited time, low motivation, gaps in understanding, and declining resilience. AI may make cheating easier, but it does not fundamentally change these underlying causes.

If we focus solely on enforcement, we risk solving the wrong problem.

When Policing Backfires

Focusing on cheating may not only be futile—it may make things worse. Efforts to ban or police AI often push its use outside school, widening equity gaps and reducing opportunities for guidance.

Many of us hoped AI detection tools would offer a solution, much like plagiarism detectors did. However, research has shown these tools to be unreliable, with experts noting they struggle to keep pace with rapidly evolving AI models (Perkins et al.). Even more concerning, they can disproportionately and inaccurately flag multilingual writers, raising serious equity issues (Zou et al.).

When detection is our primary strategy, we risk eroding trust. Students shift from learning to hiding, and we lose the chance to talk openly about how AI should—and should not—support learning.

AI Is Already the Baseline

Losing trust and guidance is perhaps my biggest concern. And that’s because AI use is already widespread and is not going anywhere.

College Board research finds that 79–84% of U.S. high school students use generative AI for schoolwork (College Board, 2025). Similarly, RAND Corporation research reports that 62% of students use AI for homework, with rapid year-over-year growth suggesting this is already an underestimate (Schwartz and Diliberti).

This is our new baseline. Students are integrating AI into every aspect of their work. If they feel the need to hide that use, they lose access to guidance—and we lose our role as partners in their learning.

Students Are Worried, Too

The good news? Students share many of our concerns.

Recent research shows that 67% of students believe AI use harms their critical thinking skills, a figure that rose significantly over 2025 (Schwartz and Diliberti). Nearly half of Gen Z also worry that AI will negatively impact their ability to think carefully (Walton Family Foundation et al.).

Students are not blindly embracing AI. They are ambivalent—and they are asking for clarity. They want guidance on when to use AI responsibly and how to navigate it ethically. This is where our role becomes essential.

The Real Risk: Missed Learning

Here is where my concern returns.

AI can support learning—but it can also replace thinking. Brookings Institution research warns that “overreliance on AI… can put children’s fundamental learning capacity at risk” and may “undermine children’s foundational development” (Burns et al.).

This is the true red flag. Day to day, it looks like students skipping drafting, bypassing reasoning, and avoiding productive struggle. These are not just academic shortcuts—they are missed opportunities to develop critical thinking.

Here is where we need to invest our energy. How do we re-engage students in the thinking process?

A Better Question for Educators

This blog series will explore how we can respond productively—by requiring visible thinking (e.g., drafts, reflections), designing tasks that prioritize reasoning over answers, and integrating AI into the learning process rather than ignoring it.

Ultimately, this requires a shift in mindset. Learning integrity is not about bans or policing—it is about protecting the learning process, promoting student agency, and fostering transparency and ethical dialogue.

So, let’s reframe our questions:

Instead of:
“How do I stop students from using AI?”
Or: “How do I detect AI use?”

Let’s ask:
“How do I design learning so AI can’t replace the thinking?”

Follow along in this series as we explore what that shift looks like in practice.

From Insight to Practice

Designing learning experiences that make thinking visible is one of the biggest opportunities—and challenges—educators face in the age of AI.

If you'd like to explore the research behind this shift, Assessment Evolved: Redefining Formative Assessment in a Generative AI Era examines how formative assessment can evolve to strengthen learning integrity, support critical thinking, and help educators respond confidently to an AI-enabled future.

Read the Assessment Evolved Report