What you see when you watch a group of eighth graders argue with an AI tool during lunch is very telling. One student asks for a new version. Another person scans an X-ray into an app to see if the machine agrees with what his doctor says. A third is figuring out how to use it to make plans for her henna business. These kids aren’t waiting for a unit on AI to be added to their curriculum. It’s already inside them, where they build things, ask questions, and sometimes yell at the screen when it gives them something wrong.
This picture isn’t always talked about when school districts and schools get together to make AI policy. The students are treated like the decision itself, not like they had a say in making it. The administrators argue. Teachers learn new things. Workshops are held by tech companies. It’s also possible that a fourteen-year-old is using AI in a way that no one in the room has thought of yet.
The issue is not with schools being smart about AI. A lot of the time, they are doing just that. Illinois just passed a law that requires classroom teachers to be involved in AI conversations across the state. This is a real and important step forward. All over the country, school districts are updating their acceptable use policies and building professional development programs to try to keep up with a trend that is changing faster than most institutions can adapt. All of that is important.
But there is a big difference in who gets to sit at the table.
Something changes when students are involved in the design process. A student who spent months playing around with AI learns how to question it as well as how to use it. When a student found out that AI could answer medical questions about his own injury, he made an app that lets him control his home computer from afar. Another person started to ask AI about how much energy it was using. These are not users who do nothing. When it comes to adults, most adults who have gone through traditional professional development programs still haven’t reached this level of functional literacy.
By looking at how students in the UK are using AI in higher education, Wonkhe found something that should get more attention than it has. What really determines whether students use AI to learn or just make things is assessment design, not AI policy. Students use AI tools in a different way when they know they will have to defend a position, explain their reasoning, or show that they understand in a real-life situation where they will be held accountable. This is not to avoid thinking at all; they use them to test ideas and arguments. Those who aren’t accountable to those further down the chain have no reason to get involved deeply. That’s still the same with the rules. The design of the task does.
There’s a chance that this is the more honest conversation—tougher to have but more helpful. It’s only the tip of the iceberg to ban AI or building detection systems. The structural question at its core is whether the assignments teachers give really make students think and what learning is supposed to look like. That problem wasn’t made by AI. It’s just made it very clear and quick.
Also, the idea that many students are building strong personal ethics around how they use these tools is something to think about. They didn’t do it because someone asked them to. Not because there is a grade scale for it. But because they are really thinking about things like effort, understanding, fairness, and even the point of learning. It’s helpful to think that way. It is more likely that schools that learn how to work with the situation than schools that are still trying to write a policy that is broad enough to cover every possible situation.
The kids have already come in. They need to know if anyone is telling them to pull up a chair.
