How to write school rules for using AI
A blanket ban does not work, because it cannot be enforced. A rule that works describes what is allowed for a given assignment, what has to be disclosed and how things proceed when there is a suspicion. Detection is one input in it, not the referee.
Concrete boundaries instead of a ban
Distinguish the types of use: research, outline, language editing, generating whole passages. For each assignment, say what is fine.
A rule that can be followed gets followed. A ban on everything gets worked around and mainly teaches students not to talk about it.
Disclosure instead of hiding
A short note at the end of the work about where AI was used resolves most disputes before they arise.
Students write it as long as it carries no punishment. That is all there is to it.
The procedure when there is a suspicion
Write down in advance what happens: a conversation about the work, questions about the content, the chance to document the process. And that the output of a tool is not on its own proof.
This paragraph protects both sides — the student from a rushed conclusion, the teacher from arguments about method.
FAQ
Should the rules name a specific tool?
Better not. Tools change faster than documents. Describe types of use.
Do we have to tell students that we check?
It is fair and practical. An announced check also has a preventive effect; a hidden one only manufactures conflict.
Who should write the rule?
Best the department together, not a single teacher. Inconsistent rules between courses are the most common source of disputes in practice.
Try it on your own text
Paste a text or upload a file and look at the score and at the specific sentences that came out suspicious.
Run a detection