Teacher urges AI literacy through controlled classroom interrogation

AI literacy – An adjunct professor describes how, in English and writing classes, students resisted “character chatbot” addiction after a structured assignment that required them to interrogate the tools instead of simply using them.

In Mike Kentz’s English and writing classes, the first sign that artificial intelligence can feel like a trap wasn’t technical. It was behavioral.

When Kentz assigned students to interrogate fictional character bots. he worried they’d end up chasing the very thing he was trying to teach them to challenge—Character.AI. which he describes as “arguably the most seductive and addictive type of chatbot for young people. ” because it offers a companion that “never says no. never gets tired and never pushes back.”.

What happened next unsettled that fear. Months later, when Kentz asked whether they wanted to interrogate another character chatbot, the reply was quick and blunt: “Nah, that’s old news.”

For Kentz. the classroom reaction answered a question that many educators are still wrestling with: what it actually means to be “AI literate.” He argues it’s not a fake idea. but “just hasn’t been fully defined yet. ” and that schools are still researching it—because literacy. in his view. depends on what students do with AI. not on warning labels.

He says the assignment didn’t create dependence. It built resistance. The exercise, he writes, helped students develop “the cognitive defense system,” teaching them to interrogate AI—one of the “best ways there is to build up” that protection.

He frames the goal in medical terms: not antibodies to panic against, but antibodies to handle exposure. “Not dependence — antibodies. ” he writes. adding that educators are still “looking for the right dosage. ” and that “finding the right balance and type of AI exposure helps the body build up its protective layers.”.

Kentz pushes beyond classroom technique to a warning he treats as urgent. AI. he says. is being sold as a product that can increase productivity and even creativity. and he argues that the only way to spell out its harms is through research—research that will “shine a light on” what he considers “its ills.”.

His comparisons are blunt and meant to land. He points to how scientists once thought cocaine was good for the body before realizing it was bad. and how cigarettes were claimed to reduce stress before cancer became undeniable. He argues that AI companies “don’t want you to know about AI’s harms. ” and that “every ruling class in history has understood that a literate population is a dangerous one.”.

In Kentz’s framing, the stakes aren’t abstract. He says the danger is like addiction to a drug—specifically heroin. cocaine. cigarettes. and alcohol—calling uneducated AI use “like getting behind the wheel with no training.” He says the concern has been with him “since ChatGPT was first released three years ago. ” and that he “often” feels it whenever OpenAI CEO Sam Altman speaks. or even just by seeing his face on his iPhone.

Yet he refuses to close the door. Turning away from the technology, he argues, is not the answer. Learning it does not mean surrender. He calls it “an act of subversion.”

To make that case, he lays out a containment approach drawn from how societies respond to drugs and viruses. “What do we do when a new drug hits the streets?” he asks. and answers with actions: contain it. capture it. make a vaccine out of it. and find a “controlled dosage” that can be given broadly once tested.

He adds that physiological resistance comes from exposure—but “not through uncontained exposure,” rather through monitored exposure. He says there are no “doctors” of AI literacy yet, “but anyone can become one.”

Kentz leans on a familiar failure in public health messaging. He says “Just Say No to Drugs” doesn’t work, and that drug literacy works instead. “The more a person knows about what is out there and what it can do to them. the more they develop discernment. ” he writes—arguing AI literacy should work the same way as vaccine development: testing it in small increments. documenting findings. toggling dosage. and combining it with other elements.

He says the broader problem is already in motion. Society is being permeated by a technology that he describes as having “never asked permission to enter our lives,” and the result is that it invades people’s systems “undetected” when adults ignore what’s happening.

His prescription is proactivity rather than avoidance. He calls it “proactivity. ” insisting it is not resignation but “strength. ” even “revolutionary.” The approach he advocates is action—“Action. not inaction”—with students and educators learning in a way that builds antibodies instead of relying on fear.

He closes with an urgent call to research and engagement. “AI is a drug,” he writes, “but if history has taught us anything, it’s that drugs need to be researched.”

Kentz identifies himself as an adjunct professor of writing at Fairleigh Dickinson University and the founder of AI Friction Labs. an educational technology platform that provides challenging. story-based simulations for educators to train and evaluate student skills. The piece appears as an opinion by Mike Kentz in The Hechinger Report. dated June 1. 2026. and is republished under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

AI literacy education classroom assignments character chatbots Student resistance Michael Kentz Fairleigh Dickinson University AI Friction Labs ChatGPT Sam Altman