Conversations about student AI use often collapse into two choices: allow it or ban it. My students’ own rules were much more specific.
They were concerned about cognitive offloading, skill erosion, loss of critical thinking, privacy, misinformation, access, and the possibility that AI could flatten the educational experience by replacing human interaction. They also worried about trust between students and educators.
The rules they wrote
- Think and try first. Ask AI second.
- Do not use AI to replace your struggles.
- Do not paste private or sensitive information.
- Double-check sources and verify the output against primary sources.
- Be honest about using it and cite its contribution when it shapes the work.
- Remain the author and the decision-maker.
- Use it to improve, not replace.
- Do the work to learn.
One student put the standard plainly: if you cannot explain the concept or the work without AI, then it is not your work or knowledge.
These were not anti-AI rules. The same students described useful applications: asking for feedback, getting a difficult topic explained in a new way, generating practice questions, identifying gaps in understanding, and using AI when they were genuinely stuck.
The distinction was not whether AI appeared in the process. It was whether the student remained accountable for the thinking.
If you cannot explain the concept or the work without AI, then it is not your work or knowledge.
“Support, not substitution” is a better starting point than a universal rule because it forces us to ask what learning the student still needs to do—and how we will know they did it.