Why Students Boo AI While Secretly Relying On It

The video is two minutes long, and it is hard to watch. A commencement speaker at the University of Central Florida tells the graduating class that the rise of artificial intelligence is the next industrial revolution. The crowd boos. She pauses, visibly startled, and tries again. More boos. “Passion,” she says, scrambling. “I love it.”

That same week, former Google CEO Eric Schmidt was hissed at during his commencement remarks at the University of Arizona for saying much the same thing. The pattern repeated at campus after campus this graduation season. Students are booing AI off the stage.

The reaction that does not match the behavior

Here is the part that should stop every educator cold. The same graduating class booing AI from the bleachers is using it constantly. According to the Lumina Foundation and Gallup 2026 State of Higher Education study, 57 percent of US college students report using AI tools in their coursework every week, and 20 percent use them every day.

So which is it? Do students hate AI, or do they depend on it? The answer is both, at the same time, and that contradiction is not hypocrisy. It is the single clearest signal we have about what AI literacy education is failing to deliver.

Maitraye Das, a computer science professor at Northeastern University who studies how Gen Z relates to AI, calls this cognitive dissonance. Students believe AI use will weaken their thinking and undercut the value of the degree they just spent four years and a great deal of money earning. They also believe they cannot afford to skip the tool, because everyone around them is using it. So they use it quietly, resent it openly, and tell almost no one the truth about either feeling.

Naming the real problem: The Confidence Gap

When a person fears a tool and relies on it at the same time, they are living inside what I call The Confidence Gap. It is the distance between how someone feels about AI and how they actually use it. On one side sits the emotion, often anxiety, resentment, or a vague sense of cheating. On the other side sits the behavior, the daily quiet reliance. The wider that gap, the worse a person feels and, importantly, the worse they use the tool.

This matters because the booing students are not making a statement about AI. They are showing us the symptom of a gap nobody helped them close. They were handed the most powerful general-purpose tool of their lifetime and given almost no framework for using it with judgment, integrity, or confidence. Of course they resent it. Resentment is what fear feels like when it has nowhere productive to go.

And the booing is the polite symptom. The quieter one is worse. Jacob Shelley, a health law professor at Western University, told Fortune he was convinced most of his class cheated on a final exam using AI. He does not entirely blame them. “I think they see through it,” he said of students who feel the system is pretending AI will not reshape their futures. When people operate inside a wide Confidence Gap, secret use and low-integrity use become the path of least resistance. Not because students lack character. Because nobody gave them a better road.

Why fear plus reliance is the worst possible combination

Consider what a wide Confidence Gap actually does to the quality of a person’s work. Someone who fears AI and uses it anyway tends to use it in the most shallow way possible. They paste in a prompt, grab the output, and close the tab fast, the way you would handle something you find slightly shameful. They do not iterate. They do not check. They do not learn the tool, because learning it would mean admitting they intend to keep using it.

This is the opposite of fluency. Research we covered recently on active versus passive AI use found that the danger to your thinking is not AI itself. It is passive, unexamined use. A wide Confidence Gap practically guarantees passive use, because shame and avoidance are not the conditions under which anyone does careful, engaged work.

So the student who boos AI and then uses it in secret is getting the worst of both worlds. They carry the emotional cost of resentment and the practical cost of using a powerful tool badly. They are not protected by their skepticism. They are handicapped by it.

Closing the gap is a skill, not a speech

Here is what will not close the Confidence Gap: another commencement speech telling students AI is inevitable. They already know it is inevitable. That is precisely why they are booing. Being told to accept something you feel powerless about does not build confidence. It deepens the resentment.

What closes the gap is the same thing that has always turned fear into competence: small, deliberate, successful repetitions. You do not argue someone out of AI anxiety. You build them out of it, one concrete win at a time. This is the entire foundation of the SeedStacking methodology, and it applies exactly here.

The mechanism is simple. Every time a person uses AI deliberately, watches it work, understands why it worked, and keeps the result, their behavior and their confidence move one notch closer together. The gap narrows. Do that ten times and the tool stops feeling like a threat and starts feeling like an instrument. Do it fifty times and you have fluency. Not because anyone gave a speech. Because the person stacked enough small wins that the fear ran out of room.

Want to find your own Confidence Gap? Download the free Confidence Gap Self-Audit worksheet, a 15-minute exercise that locates your gap and maps your first three closing wins.

For educators, this reframes the whole job. The student using AI in secret is not a discipline case. They are a person with a wide Confidence Gap and no one helping them close it. The most useful thing you can do is not catch them. It is bring the behavior into the open and give it structure: here is how to use this tool well, here is how to check it, here is how to keep your own judgment in the loop. That is not permission to cheat. It is the only known cure for the quiet, low-integrity use that fear produces.

Want a structured way to do this with your students or your team? Join the free Harvest Kernel community

The graduates are not the problem

It is easy to watch the booing videos and see a generation that is difficult, contradictory, or ungrateful. That read is wrong, and it is lazy. The class of 2026 is the first cohort to move through a full degree with these tools in hand and almost no one teaching them how to hold them. They are not contradictory. They are uninstructed.

The boos are a request. They are asking, in the only language available to a crowd, for someone to close the gap between how AI makes them feel and how they are actually living with it. That is a teaching problem. It has a teaching solution. And it starts the moment we stop telling people to accept AI and start showing them, one small win at a time, how to be genuinely good at it.