Why AI Bans Don’t Work

AI Bans Feel Safe. But We Are Leaving Students Behind.

by Jim Bowler, engAIged

Artificial intelligence is creating a new kind of anxiety in schools. Parents are asking tough questions. Teachers are trying to figure out what counts as appropriate use. School leaders are wrestling with privacy, cheating, screen time, bias, and tools that seem to change every few weeks.

So it’s understandable that some districts are reaching for the simplest response: let’s ban AI. I understand the instinct. But that doesn’t serve students well, and that should be our main concern.

A Ban Is a Rule. It Isn’t a Strategy 

There are legitimate reasons to be cautious. A six-year-old shouldn’t have unrestricted access to an AI chatbot. Students shouldn’t be entering private information into tools nobody has vetted. Teachers need clear expectations about when AI assistance crosses into doing a student’s work for them.

Those are real issues. They require immediate guardrails.

But guardrails and bans are different things.

New York City just announced a one-year moratorium on generative AI for students through eighth grade — the most expansive ban of its kind in the country. Elsewhere, groups of parents and educators in some districts are calling for the same thing, or even a complete ban.

For me, a ban means we don’t know what to do. It postpones the decision rather than making one.

Students Don’t Get Those Years Back 

School systems move slowly. We know that. Students don’t.

A district can spend a year studying AI, another year developing policy, and another on professional development. A teacher may still be there when that process wraps up. A high school junior or senior won’t be.

That student graduates into a world where AI is becoming part of writing, research, coding, finance and business. There’s something ironic about telling students AI will transform nearly every career they’re considering, while telling them they can’t learn to use it at school.

Our job isn’t just to protect students from the world they’re entering. It’s to prepare them for it.

We’ve Had This Debate Before 

Calculators were once criticized because students might stop learning math. The internet raised fears about unreliable information and shortcuts. Those concerns weren’t imaginary.

Districts didn’t solve the problem by banning the technology. We learned to distinguish between when it supports learning and when it replaces it.

AI requires the same judgment. There will be plenty of times students should write, calculate, or research without it. Teaching them to know the difference may be one of the most important skills we give them.

What Guided Use Can Actually Look Like 

A high school study this past year found that students demonstrated greater autonomy when they worked collaboratively to evaluate AI-generated writing (Riser, 2025). A separate review of AI tutoring studies found positive effects on K-12 learning and performance (Létourneau, 2025). In both cases, however, AI was part of the learning process rather than a replacement for teaching. Teachers still played a critical role in accountability, discussion, and judgment.

The research doesn’t suggest that giving students AI automatically improves learning. It shows something else: guided, critical engagement with AI can build exactly the judgment and autonomy schools say they want students to have.

I’ve seen this in simpler ways, too.

I recently suggested to a sports journalism teacher that students use AI to “interview” an athlete, past or present, whom they admired. Students conducted the interviews, shared their transcripts, and debated whether the athlete had really been represented accurately. It became one of the class’s favorite activities, and it sparked exactly the kind of critical discussion teachers want.

Contrast that with feedback I received from an AI skeptic: “If I can’t bring the real person into my classroom, we won’t do the interview.” I’m sorry to say that Abraham Lincoln or Rosa Parks aren’t available for a visit.

AI can create learning experiences that weren’t possible before. Our job is teaching students to approach them critically — which is a very different job than teaching them to avoid them.

Guardrails Now. Better Guardrails Tomorrow 

If I were leading a school today, I wouldn’t pretend we have all the answers. We don’t.

I’d build the best guardrails I could: protect student information, define appropriate use, differentiate by age, train teachers, communicate with parents, and teach students to question AI rather than trust it by default. Then I’d keep studying what works and adjust it regularly as we learn.

This approach requires more judgment than a ban. But our students need something better: thoughtful preparation for a world that is still taking shape.

The goal shouldn’t be to ban AI because its future is uncertain. This is precisely the time to teach students and educators how to use it wisely.

Reader Question

What’s your school’s actual AI policy? Ban, free-for-all, or something in between?

---

How can you get started using AI to save you time and improve learning in your classrooms? Download our guide free.

---