** Work in progress **
If you have suggestions, please pass them along to Steve Jung ([email protected])
AI Talks - What to Say When a Student Denies Using AI
If a student is denying using AI it is probably for one of three reasons; 1, they honestly didn't use AI, 2, they used AI, but don't understand that they did, or 3, they did use it, but they don't want to admit it. As the faculty member, it is your struggle to figure out which of the three is true.
Of the three options, the second, use, but not understanding, is the easiest one to deal with. In this case, you just need to instruct the student as to what AI is and how they may have misused it. Here might be a good time to explain that there is a difference between use and misuse on AI. Use of an AI for brainstorming is probably OK; they can do that with a friend from class. Misuse of AI is having the AI write part of the paper or having the AI do the "research" for the student.
The first option, that they didn't use AI, this is probably the hardest to deal with. The reason this is the hardest is because you probably have some AI detectors saying this appears to AI or you gut tells you this isn't written by a student. AI writing is currently difficult to detect and it is getting harder every week. In the beginning of AI writing it wasn't too hard to guess if something was AI. AI writing was boring, plain, generic, soulless, repetitive, not creative, perfect spelling and grammar, but now AIs can rewrite that original perfect, but bad writing, and make it readable and more "human". Students can paste that first draft into the same AI and ask it to improve the writing or they can use the "humanizer" AIs to make it sound more human. Students that are falsely accused are flagged for a few reasons: 1, they used some sort of translation software, 2, they accepted the recommendations from Grammarly, Microsoft Word, or Google Docs with Duet (all now have AI agents built-in), 3, they wrote in a very mechanical manner (heavily formatted and stylized lab reports or case studies) or traditional 5-paragraph or 7-paragraph essays. This means that English language learners, STEM students, and students on the Autism spectrum will be falsely flagged more than they should be.
In regards to the third option, the student denies using an AI, but you suspect that they did. In this case, you either need to "prove" they did or get a confession from the student. Honestly, without taking their computer and checking their browser history, revision history of the paper, and computer cache you have no way to "prove" that the student used AI, the AI detectors don't prove anything, they just analyze probability. So, the only real option is to get a confession out of the student. [** news on this front - Grammarly for Education is starting Grammarly Authorship which will now track the writing process. It will track cut / paste operations from various websites or from other programs. At the end of the process will be a "report" on original words typed, words pasted from AI, words edited in the AI section, and words pasted from other sources. This currently in Beta and should become live in early 2025. Maybe we should require students to submit the report with their assignment?]
Trying to Get a Confession
Trying to get a confession requires various appeals to the student's integrity. You can appeal to their personal integrity and character, appeal to their future work and income, or appeal to their fear of punishment.
We most often appeal to punishment. It usually works. We express to the student that they can lose points, a grade for a paper, fail a course, and even lose scholarships that are academically based. We also threaten with the report of the incident to the VPAA and that this will go on their permanent record. All of that is fine and good, but to make it stick we need proof or a confession. If this appeal doesn't work, we have neither. Do not start with this, it is not the most effective AND it has the possibility of getting the student to stop talking because they know that you have no proof.
Another appeal, and one that you might want to start with, would be to appeal to their future. A simple violation that goes to the VPAA will remain on their permanent record and more. That mark will probably ruin their chance to get into grad school or end some chances for letters of recommendation for work. Right now there are university presidents being forced to resign because of accusations, with some proofs, of academic integrity violations from their earlier academic career (plagiarism and falsifying data). Company founders and CEOs losing their jobs, careers, and futures because of integrity violations. Authors and academics nearing the end of their career have their entire career destroyed over simple things like failing to cite sources and other violations of academic integrity. There is a database that was created just to track retractions by authors/academics (link: http://retractiondatabase.org). AI misuse is just as bad as any of those other types of violation and they all end badly.
This appeal to their future work, career, and reputation will often mean very little to the student. We live at a time when people are living for the moment and looking for the quick and easy. The long term perspective is missing for most young people; they barely see past the end of the semester, let alone their degree program. So, this appeal, though it should hold weight, will often just cause a pause. Again, the student knows that without proof, you are relying on a confession. They know that if they can wait out your "lecture", they will probably get away with it. That doesn't mean that we should not use this appeal. This is just one part of the approach to get a confession. Starting with this gets them thinking.
** Still working on this **
One major appeal is to their character; do they want to be known as a cheater or as an honest and upright person. At an institute of Christian higher ed, this ought to be effective, but it depends on the character of the student.
We want our students to become mature Christians and "impact the world for Christ" (mission statement). To do that we need to instill, with the help of the Holy Spirit, integrity. Let them know that once a paper comes back as plagiarized or AI, that student will be viewed, unfortunately, as a cheater. The point being that lying now, can cause problems for a long time and will not help them in the long run. They need a reminder that honesty and integrity are better than a grade; that's a life lesson.
A good name is more desirable than great riches;
to be esteemed better than silver or gold.
Proverbs 22.1
** Also working on **
Another aspect that we need to teach the students is that they are earning a grade; they are earning a degree. We want the work of the student to be the work of the student. We want to hear the student voice. That voice, they, as people, gets lost when the AI does the writing. The writing may be good, but it is not the student's voice. We are instructing students, assessing student learning, and giving the student a grade. If the work is from AI, it is not from the student.
That concept is what we need to instill; we want to educate the student and accurately evaluate the student. And this leads to questioning the student; is what you turned in your work? the work of an AI? or your work augmented / supplemented / complimented by AI?