When a Student Turns in AI Writing: What to Do in the First 24 Hours

Teachers know their students, so it’s easy to notice when the vocabulary in any given essay is too polished. When the sentences are all the same length and the ideas are right but use a lot of words without saying much at all, it’s natural to suspect they turned in AI writing. But what do you do with those suspicions?

Moving quickly matters, so here’s what teachers should do in the first 24 hours after a student submits AI work:

Think It Through

It’s unfair for step one to be “make an accusation,” so resist the urge to fire off an email or mark a zero on the spot. You’ll regret making a rushed judgment call before you even have a conversation with the student.

Instead, pause and reexamine the student’s work with a specific question in mind: What, exactly, made you think they used AI to complete this piece? A vague sense of unease isn’t enough, but documented specifics are actionable.

Document What Made You Suspect AI Writing

There are dozens of AI detection tools, and many of them are free — your school or district might already have one you can use. But these tools aren’t perfect and false positives are common, especially when judging writing by non-native English speakers, so your own judgment is more reliable. Read the work again and make note of:

• Sentence rhythm that never varies. Human writers naturally mix short lines with longer ones, but AI text tends to repeat the same length and format sentence after sentence.
• Hedging. Frequent use of “might,” “could,” and “arguably” can be signs of a tool that’s been trained to sound balanced rather than a student taking a position.
• Polish without meaning. Technically clean and grammatically correct writing that says nothing personal, makes no real argument, and takes no risks is classic AI.
• A sudden jump in quality. If the errors this student usually makes have vanished overnight, that’s worth flagging.
• A departure from your student’s voice. You know your student, and you’ve read their work before. Does this sound like them, or like someone else entirely?

Alone, none of these would be enough, but if you’re already suspicious and you can point to multiple examples of these in the work submitted, it’s a reason to talk.

Have a (Hard) Conversation

Academic dishonesty and integrity are serious issues, and bringing them up with students (and their parents) is never easy. However, these conversations are opportunities for you to teach and for students to grow, so they’re worth getting right.

Don’t approach it like an interrogation. Instead, use curiosity as a starting point and open with something like, “Walk me through how you wrote this” or “Tell me about your thinking on this paragraph.” These open-ended invitations will give you more to work with than simply asking “Did you use AI?”

However, when it’s time to turn the conversation to AI, reassure the student and ask about their process, including how they got from the prompt to the finished piece. Given the chance to answer a genuinely open question, most students will tell you what happened, and you’ll rarely need to prove anything if you just ask.

Two Ways the Conversation Can Go

If the student used AI, no version of “Why did you cheat?” is going to be a useful question. It’s important to identify the factors that led the student to use AI, like time constraints, confusion about the assignment, or just not knowing where to start. You could begin by asking, “What stopped you from completing the assignment on your own?”

Based on the student’s answer, you could plan supports that might help them in the future, and you could even talk with them about how to ask you for help sooner next time. For the current assignment, agree on what happens next. Options may include rewriting independently and resubmitting; rewriting during class; or something else based on your school or district’s policy. No matter the next steps, be clear about what you expect going forward.

If the student didn’t use AI, thank them for talking it through and providing clarity, then let it go. There’s no use in letting a suspicion linger, and a student who feels wrongly accused will remember it longer than you will.

Either way, loop in a department head or administrator if your school has a policy that calls for it. Sharing a quick note now is easier than trying to remember the conversation later.

The Longer Fix Past 24 Hours

The teachers who worry least about students using AI have usually redesigned how they assign writing in the first place. They build more time for in-class drafting and more one-on-one conferencing while students write. You can also make documenting the drafting process part of the assignment, with scores for an outline, action taken on initial feedback, and revision history, not just the finished product. When you’ve watched a piece of writing take shape in front of you, there’s less room to wonder how it was made.

Making a shift like this also means teaching AI literacy directly rather than pretending the tools don’t exist and aren’t already in students’ lives. Giving them guidance about when it’s appropriate to use AI, when it isn’t, and how to evaluate the outputs tools produce will help you get ahead of the curve in the short-term and set them up for responsible usage in the long-term. Ideally, it will also give you fewer cases of AI suspicion to investigate as students who already know where AI falls short choose to do the work themselves.

Minimizing Your Teaching Stressors

When you handle a case of suspected AI use with curiosity and good faith, it’s rarely the disaster it initially feels like at 9 p.m. while you’re staring at one suspicious paragraph.

As a teacher, you already deal with unique demands that can leave you feeling drained. With the right approach to conversations about AI, you’ll be able to minimize some of that stress. For more strategies to set you up for a long, fulfilling career in education, download our workbook From Burnout to Balance: Protecting Teacher Well-Being.