Barbara S. Held is the Barry N. Wish professor of psychology and social studies emerita at Bowdoin College.

Now that it’s end-of-term grading season, college professor pals say I was wise to retire from teaching before AI infected students’ work, making grading more onerous than ever.

AI-driven cheating has become so ubiquitous that some professors are turning to “old-school methods” like bluebooks and oral exams, while others are fighting fire with fire by using AI deviously to catch cheaters.

As history professor Will Teague wrote about his use of the “Trojan horse” trick in his Huffington Post piece titled “I Set A Trap To Catch My Students Cheating With AI: The Results Were Shocking,” “I inserted hidden text into an assignment’s direction that the students couldn’t see but that ChatGPT can.” 

The assignment required students to describe key points about “a thwarted rebellion of enslaved people in 1800” made in a reading assignment. Teague doesn’t disclose the hidden words that sent AI cheaters in the wrong direction, but they can be anything that has that effect. 

According to higher education influencer Derek Newton of “Cheat Sheet,” some educators “find this Trojan horse approach to be unethical, a form of gotcha.” They may see it as entrapment by enticing students to behave in ways they wouldn’t behave normally.

Yet Newton says he’s fine with it: “If you do the work of learning, you have nothing to worry about.” I agree with Newton. After all, professors have been using different methods to detect cheating for centuries, and I don’t know of any case where they were considered unethical for doing so. So why does this form of cheating detection cast professors as immoral in the eyes of some?

Case in point: The moral compass of a professor pal of mine is beyond reproach. Lying, cheating and manipulating others are just not in his nature. Having taught for decades, he’s never been prone to suspecting the worst of his students. But now being bombarded with AI-generated papers, even he has resorted to a sneaky form of detection that I find ingenious.

In one example, students were assigned a short paper in which they were expressly told in class to describe the two approaches that we all use to attribute beliefs and desires to other people, according to an assigned reading and in-class discussion. He worded the assignment’s prompt such that students who used AI would be led to discuss psychoanalysis, which had nothing to do with the assignment: “When Sigmund sees Dora enter his office and lay down on his couch, he attributes beliefs and desires to her to explain her behavior. How might he be doing this? Discuss two options.”

Note in this prompt his use of the words “Sigmund” and “Dora” (see Freud’s famous case of Dora) as well as reference to the couch, which serve as distractors designed to mislead AI-using students.

Sure enough, several students turned in papers about psychoanalysis. The beauty of his method is that it requires no tech savvy. Moreover, the professor doesn’t have to confront students on their use of AI, because their grade of “F” is based solely on their failure to address the assignment, which is how he explains that grade to those students respectfully.

And he says he’s proud of himself for conjuring up this form of AI-use cheating detection, as it indicates he’s finally developing certain self-preservation skills, which he has long felt he’s needed to cope with various life events.

I give three cheers to teachers at any level who hone their cheat-detection skills. They should not be considered immoral for doing their job, which is to educate. Because students who cheat are thwarting their own educations, being caught cheating should be looked upon as a student’s opportunity to become a better learner — indeed, a better person — as long as the teacher explains the problem to the student respectfully. 

With President Donald Trump moving to dismantle the Education Department, we need more educational justice in our country, not less. Celeste Fernandez put it in the title of her National Education Association piece, “Trump administration robs students of educational opportunity with annoucement to dismantle Department of Education.” If using AI to detect cheating helps teachers strengthen educational justice in that respect, so much the better.