** Work in progress **
If you have suggestions, please pass them along to Steve Jung ([email protected])
AI Talks - What to Say When a Student is Suspected of Using AI
- Do you have a guess as to why we are having this conversation? [And then let them stew. Wait for them to make a guess. This will be uncomfortable for both of you, but it allows them an opportunity to confess before it gets too accusatory and they get defensive.]
- Your writing set off some AI detectors. [Don't tell the percentages because those are just numbers based on probability and not very accurate. The important thing is that some AI detector notes that there is a high likelihood that part of the paper is AI generated.]
- Can you tell me why you think that could have happened? [Again, wait. Give them the opportunity to "confess" before you get into the act of shaming them for violating academic integrity. They may have used an AI for some small part or maybe did something like had it rewrite a sentence or two. No matter what they did, you still need to judge if it was truly a violation of academic integrity.]
- Did you use any translation apps, like Google Translate? Those can produce false positives. [Acknowledge that there are false positives and that this really could be a mistake. You are investigating, this is not a trial; you are trying to determine the best course of action and potential punishment based on the facts. And, yes, translations from other languages are typically flagged. The most common translation is of one word, or phrase, to another, and that matches how an AI writes.]
- Can you open your paper on your computer? I'd like to see if we can go over your "revision history". I am just trying to see where and when an AI portion may have entered your work. [If the paper doesn't have a history, then ask why. That might be a big clue that they just copied and pasted from an AI. But, if the student has a lot of writing and revisions you can know that they wrote, but may have used the AI to "improve" what they wrote. Use your best logic and intuition to "read" the revision history. ** to learn how to open the revision history: support.microsoft.com]
- How about any notes? Do you have any notes or articles that might have your writing on them? [If the student doesn't have notes or resources, then you should be really suspicious about their paper. If they can produce these things and you can see the relationship to the essay/project, then you may have to dig deeper into when and where they used they AI.]
- I am just trying to see where the AI detector might have flagged something. This is as much a learning exercise for me as it is for you. [Assure them that you are not here to accuse them, but to learn why the AI detector may have flagged their paper. If they are guilty, confession and contrition are great responses and learning experiences.]
- Did you happen to "correct" your paper with Grammarly? With Microsoft Word? Google Docs with Duet? Those all have AI built into their processes. That may cause some flags as well. [Again, wait. Allow them to describe how they "corrected" their paper. Ask probing questions. It is one thing to have a word misspelled and corrected by the AI. It is an entirely other thing to accept the AI's revision of a sentence or paragraph. Misspellings corrected will not cause a detector to flag a paper. Rewording sentences by the AI will cause flags.]
- Did you use an AI for brainstorming? Possibly for taking notes or summarizing articles? If you pasted those into your paper, rather than putting it into your own words, that could have raised some flags too. [Again giving them an opportunity to explain how and where they used an AI. And if they did use the AI as a resource (summarizing or writing), then it needs to be listed as a source or co-author.]
- Did you use a "writing" website to improve your writing? If so, which one(s)? [There are AIs that "humanize" other AI writing. They change a generic AI essay into something that sounds more human (still perfect grammar and spelling, but more exciting). Explain to them that those websites are most likely AIs that are changing what they wrote. Whenever that happens, it is no longer the work of the student, but something "written" by AI and that is a violation of academic integrity. Qullbot is an AI that rewrites entire paragraphs, there are others.]
