In Ohio, there is at least one classroom where the teacher doesn’t ask if a student used AI. Now the question is how they used it and what they added that wasn’t in the prompt. It sounds like a small change, but it is making people rethink what academic integrity even means in the year 2025.
Kritik is an edtech platform that just released VisibleAI, a tool that keeps track of the writing process in real time. It keeps track of changes made, highlights passages that were helped by AI, and shows teachers not only what a student turned in but also how they got there. There is a record of each change. Every time AI does something, it leaves a trace. A lot of teachers have been waiting for this kind of openness, but some students may find it a little unsettling.
It’s possible that this is exactly what college needed. For many years, detection software—platforms that gave a grade based on a percentage and let professors make decisions based on little evidence—was the main response to AI writing tools. Those tools didn’t always work right and often wrongly flagged people who didn’t speak English as their first language. VisibleAI does things in a completely different way. It doesn’t ask if AI was used; instead, it tries to show how much original thinking went into that use. No, that’s a very different question.
In a recent panel discussion, Dr. Keely Croxton, Associate Dean at Ohio State’s Fisher College of Business, put it in a clear way. She said that the point is not to catch students, but to see what they’re thinking. If a student takes a draft made by AI and changes it using real judgment and insight, that shows you something real about their skills. The last document doesn’t do it by itself.
Dr. Andrew Reffett of Miami University talked about one method his teachers use: they give students a printout of an AI-generated answer and ask them to comment on it, make it better, and explain what’s missing. You don’t have to make something from scratch for the test. It means to think critically about something that already exists. That’s more like what most professionals do with AI at work: they test it, criticize it, and make it better.
But it seems like these kinds of tools are coming into a world that isn’t quite ready for them yet. A lot of students still don’t really understand what it means to use AI in a responsible way. The limits change for each class, teacher, and school. Now there’s a system that keeps track of all the changes they make. With the best of intentions, that might not feel like guidance at all and more like being watched.
The worry about an arms race is real. Tools for finding things have always been behind. Students found ways to get around Turnitin. If they’re determined enough, they’ll also find ways to get around this. The more interesting question is whether this level of openness will help or stress out the majority of students—the ones who aren’t trying to cheat but are just confused about the rules.
VisibleAI seems to bet that being open about things will change how people act. If students know that other people can see the changes they’ve made, they might think more about how they’re using AI. They might learn what teachers now call “AI literacy,” which means knowing what these tools can and can’t do in a useful and critical way. There is research that shows that using AI feedback along with structured reflection makes writing skills better, not worse. That’s worth some money.
Still, the argument for durable skills like critical thinking, adaptability, and original judgment only works if technology is used to improve those skills and not just look for their lack. A tool that keeps track of bad behavior without teaching students how to behave differently is just a fancy version of the old model of punishment. It’s likely the way teachers use it that makes the difference.
