Raising a concern about AI use is uncomfortable for both of you, and how the first conversation goes often shapes everything after it. This guide helps teachers prepare, ask open questions, look at drafts together and reach an outcome that feels fair to everyone involved.
Key takeaways
- Gather your evidence and re-read your policy before you meet.
- Start with open questions about the work, not with the score.
- Drafts, notes and version history are the fairest evidence for both sides.
- Leave room for honest mistakes and unclear rules, and agree on next steps.
Before the conversation
Take a little time to get clear on what you actually know, and what you don't. A few minutes of preparation makes it much easier to stay calm and specific once you're face to face.
- Re-read your policy. What does your school or syllabus say about AI tools? Was the expectation for this assignment clear to students?
- Write down specific concerns. Quote the passages that worried you and say why: a change in voice, a source you couldn't find, highlights from a scan.
- Collect comparison points. In-class writing, earlier assignments, and any drafts or document history you can access.
- Decide the goal. For most first conversations, the goal is to understand what happened, not to deliver a verdict.
Opening the conversation
Choose a private setting and a calm moment, not the end of a lesson in front of classmates. Keep your tone neutral. A simple opener works well: 'I'd like to hear how you put this essay together. Can you walk me through it?' Avoid words like 'cheating' or 'caught' at this stage. They shut a conversation down before it has started.
Then listen. Students who wrote their own work usually talk about it easily, even when they're nervous. They remember where an idea came from, which source was hard to find, which paragraph they rewrote three times. Silence or vague answers aren't proof of anything either, so give them time and come back to a question if needed.
Questions that open things up
- How did you choose your topic or thesis?
- Which source was most useful, and what did you take from it?
- Can you explain this paragraph in your own words?
- What was the hardest part to write?
- Did you use any tools while writing, like a grammar checker, a translator or a chatbot? How did you use them?
That last question matters. Many students use tools in ways they don't think of as AI use, and many policies are vague about what's allowed. Asking directly and without judgment often gets an honest answer. If a student says they used a chatbot for ideas or to fix their grammar, ask them to show you what that looked like. Their answer usually makes it clear whether it crossed your line.
Looking at drafts and version history together
If the work was written in a shared document, version history can show it growing over time: notes turning into paragraphs, sections rewritten, comments resolved. A long essay that appears in one paste is worth asking about, though there are innocent explanations, like drafting offline and copying it in. Ask the student to show you their process rather than demanding proof.
If you scanned the work with the AI detector, you can look at the highlighted sentences together and ask the student to talk about those parts specifically. It keeps the conversation concrete and gives them a fair chance to respond. If they disagree with a highlight, hear them out. They may well be right.
After the conversation
- Write a short, factual note of what you discussed and what evidence you saw.
- If you're satisfied the work is the student's own, say so clearly, thank them for their patience and move on without lingering doubt in your grading.
- If concerns remain, follow your school's process and tell the student what happens next.
- Where the rules were unclear, consider offering a chance to redo the work, and clarify expectations for the whole class.
For the signs worth noting before a conversation like this, see how to tell if an essay was written by ChatGPT. Our page for teachers covers a complete review workflow.