A.I. Doesn’t Cause Cheating. Fear Does.


Front Royal, VA. How do you catch a cheater?

This is the question that is plaguing the minds of college and high school faculty across the country this spring. No longer can we rely on our own good sense and knowledge of the literature or the help of tools like Turnitin.com to detect academic dishonesty in student papers. No, now we must find ways to detect a much more sophisticated kind of cheating: the quite convincing fake papers written not by students themselves or by someone out on the internet but by artificial intelligence.

This new development is, of course, very important to consider. But in attending to the means of cheating I worry that we may lose sight of an even more important question. Because while A.I. makes cheating easier and more effective, A.I. does not cause cheating. So what does?

What makes students cheat?

As a sometime history professor, dealing with cases of academic dishonesty has been one of the most unpleasant and gnarly problems I have faced in my teaching. Even though I have taught mostly honorable students who take the Ten Commandments seriously, there have also always been those young people who give into temptation. And now that A.I. has reached the sophistication needed to create at least C-level papers that are by-and-large indistinguishable from honest C-level work, we can expect the problem to grow significantly. Sloppiness, inconsistency across assignments, and the appearance of illogical adjectives (a sign that a student has copied material and just switched out some words for synonyms) will no longer be enough to raise teachers’ suspicions.

My husband, a full-time professor, notes that there are two kinds of cheating student: the basically moral person who gives in once to temptation and, when confronted, confesses and apologizes; and, far harder to handle, the thoroughly dishonest and committed person who, even when confronted with incontrovertible evidence of his or her cheating, will deny the offense to the death (even if it means letting another student take the blame).

Both situations are shameful and distressing, but the second far more so. And in both cases, students knew better than to cheat. Yet they did it anyways. Why?

One probable reason that comes to mind is the way that student cultures treat academic dishonesty. As a group, students at any particular institution often have a very different idea of what constitutes cheating than do teachers or professors, even if the rules are made clear by institutional policies. At my high school in the late 1990’s, for example, copying homework was considered cheating by teachers but not by students. Most “good” students frequently divided up homework tasks and then copied each other. We considered it clever time management when facing our heavy homework loads. And many, many students cheated on exams at one time or another, including, according to my personal knowledge, all but one of the top ten students in my graduating class. Sometimes this cheating was premeditated and thorough. Other times it took the form of passing a note to a seatmate in a fit of desperation. It is relatively easy for students to convince themselves that this is not real dishonesty when such strategies become widespread in a student culture.

In terms of college cheating and student culture, I often wonder at the stories I hear. In this era of retweeting and reposting material, do students just no longer take it for granted that sources should be cited and that proofreading a friend’s essay should stop short of rewriting? Anecdotal evidence suggests that many students really don’t have an accurate common sense about these matters. In spite of coaching by professors, students often edit each other’s papers in a way that would give me pause but does not seem to bother the students. Like my high school classmates, they just don’t think that it’s cheating when they collaborate on papers.

But–shouldn’t they know that it’s not quite honest? Do they really not get it? Because, you see, in high school, we actually did know that copying work might not be considered completely honest by our teachers. We, the students, all told ourselves and each other that it was fine, but we knew it wasn’t. And I’m sure that students who passed notes during tests knew that that wasn’t really okay, either, even though it wasn’t quite the same as bringing in a premade cheat sheet.

So I suspect that today’s students also really do know the truth about at least some of their casual academic dishonesty. Otherwise, neither of the two categories of cheaters that my husband mentions would exist: neither the contrite ones nor the liars. For both realize very plainly that they are in trouble. And both know why. They are not, in short, usually surprised.

When I teach college courses, my academic dishonesty policies are strict, even severe. Partly this is because I know that it did me no good whatsoever to participate in a culture of copying homework in high school, however casually. I know that as high as the stakes may seem to a cheating student, it is better for their dishonesty to be caught and reckoned with at age 18 than at 35. And I know that the moral harm that cheating does to the soul and mind is much greater than the harm done by a “zero” on an exam and by the shame and disappointment felt by the cheating student and his or her family.

Yet I also know that I probably do not actually catch most of the (few) students who cheat in my classes. And nowadays when a cheating student no longer even has to go to the trouble to steal another student’s paper, or pay a dishonest colleague to write a paper for him, or even just search the internet to copy something from Wikipedia, but can just pay for an AI-written paper that will likely evade detection; well, let’s just say that I expect even more undiscovered dishonesty to come.

And that is a darned shame, and I hope we can figure out a way to do something about it. But for whom is it a shame? Not for me, the professor, not really. And not for the other students–again, not really. As we all know, the real harm comes to the cheater.

Because unlike it was for one high school classmate of mine who copied an essay from an obscure magazine and then turned it in to a teacher who happened to be a subscriber, today’s new brand of cheater, the one who uses Chat GPT or whatever the next new technology will be, will not get caught. And so the price she pays will be far greater. She will walk around well into her adult life knowing that people only approve of her because she lied. And that is what we ought to be worrying about.

A world populated by the heartbroken and insecure, by those riddled with imposter syndrome, is not one in which I wish to live. Do you?

Which brings me to the second reason behind cheating at my high school. That reason is, simply put, fear. Fear that if you did not perform, something terrible would happen.

I think we can all agree that this kind of fear is behind most cheating. Yet there is little that can be done about it on an institutional scale. Strict or lenient policies about academic dishonesty do not, on their own, affect the interior fear, borne by most cheating students, that unless they cheat, they simply will not be good enough for their parents/peers/job applications, etc. This is the same fear that leads to A- students freaking out, for lack of a better term, over a perfectly good 92% because it is not an A. A 92% (or an 85%, for that matter!) is a good grade, kids! You’ve done just fine.

These students cannot accept that being good enough is in fact good enough. They must be—always and everywhere and in every way—excellent. I suspect, to my great distress, that their very self-worth lies in achieving this level of performance. And isn’t cheating a type of performance, after all? The performance of a lie?

Rather than obsessing about how to catch all forms of cheating, we need to consider whence this fear of not being excellent comes and how we might affect it. Why do students turn in stolen papers instead of turning in two terrible last-minute pages of their own? Why run the risk of getting a zero when, in fact, an honest but horrible single paragraph might get them thirty out of a hundred points? Professors aren’t monsters—we hate grading terrible papers, but we much prefer them to dishonest ones.

This is a problem to which I do not know the answer. In an ideal world, academic departments would each take a three-day retreat to discuss fear and student culture and then devote a week of classes at the beginning of the academic year to addressing this fully with students. We could read deeply about fear and psychology and have presentations on psychological studies of cheating.

This is, alas, completely unrealistic. Faculty at many schools barely have time to meet as a department at all, much less to study this question together in an organized way. Instead, we will need to think about and discuss this broadly over time. As a beginning, I would like to offer three small-scale suggestions.

First, while teachers have limited influence over highly peer-oriented student culture, we can give students better information about academic honesty. We can take a little extra time in the first class meeting of each semester not just to go over our cheating policies but to have a discussion of a variety of detailed scenarios to see what students think of each of them. Is having a friend correct grammar in your essay honest? Is having a friend suggest how to rewrite a sentence? What about working through an entire paragraph together? And what about using Google docs so several friends can read and edit it for you? And what is a paraphrase, really–how much of the wording has to be changed? Is it “yours” if you just change out some synonyms?

When dishonesty is a miasma, it’s too easy for students to let themselves off the hook. Giving students specifics in a low-stakes, conversational setting might reduce the potential for honest “I didn’t know” mistakes and might also help students be more forthright with each other in private situations that skirt the line.

Second, we can help students put their perfectionism and fear of failure in perspective by using tips learned from anxiety research and cognitive-behavioral therapy. I like to call this the “And then what?” approach. In the academic dishonesty spiel in the first class of the semester, why not ask students to talk through a couple of scenarios (make a chart or list on the board) with you to figure out what would realistically happen if they turned in one terrible page on a paper assignment and what might happen if they turned in something dishonest. Lead them through by asking “and then what would happen?” to show that in the end, one bad grade on a paper is not in fact disastrous—they will recover—but getting a falsely earned B because of cheating causes personal and social harm that will affect them in the long term. Be completely frank.

In other words, since telling students that grades and performance don’t really matter would be disingenuous and unconvincing, perhaps we can we encourage them to think about the things that actually matter more. It might also be worth reading out the G.K. Chesterton quote about how “anything worth doing is worth doing badly” and asking them how they think this applies to study, although we’d have to be careful to emphasize that we’re not encouraging sloppiness or laziness.

Finally, we might take a little time to consider some additional individualized approaches to keep in our toolboxes for when a particular student seems to be struggling with fear and a sense of self-worth based on performance. This is for the A student who is dismayed by an A- and comes to office hours to challenge your grading, or for the student who is frustrated with regular Bs on papers and exams but does well in discussion, or vice versa. What are some creative experiments we might try?

For example, while grade-based assessment is too useful to abandon overall, what about asking a generally well-performing but anxious student if he or she would like you to stop putting grades on his or her exams or papers for the remainder of the semester, recording them privately, instead? What does the student think might change if they were to do this? You can promise to tell them if they’ve dropped below a B+ or so, if needed. I have seen this have powerful and long-lasting effects on learning priorities and a personal sense of self-worth.

Alternatively, students who do well in discussion but not in writing or exams might appreciate your help in setting personal writing or study deadlines for themselves. Of course, we already suggest to students that they give themselves deadlines, but perhaps you could be the keeper of the deadlines for them. One of my college classmates, for example, had her goals pinned onto her thesis advisor’s wall: “No horseback riding unless there’s a full draft by April1!”

This is helpful for the kind of student who develops hives when she does poorly on the objective section of an exam because she thinks of herself as “a good student,” or the young man who can’t face accepting that he needs to start working on his papers earlier and so deflects by deciding that you are just an unfair grader. A little extra step-by-step accountability can give students needed tools here.

And then of course, when a student does cheat, it is best for them to be caught, and so we do need to continue considering how to detect academic dishonesty. The zero, the shame, and any other reasonable penalty is much better to accept than it is to go undetected and so continue cheating and continue believing that you cannot survive disappointing their teachers, their family, and themselves.

The foundational issue behind cheating is not figuring out how to stay one step ahead of the newest cheating technologies, but rather finding ways to convince students that honesty is actually safer for their souls than dishonest success. And penalties, though useful, will only go so far in the face of the greater fear of not performing.

For we will not improve our students’ honesty just by figuring out a way to identify AI-written essays. They will just continue to find better and better ways to cheat. Improvements in honesty will only come when we can give our students tools to overcome their crippling fear of being average.

Image Credit: Thomas Malton, Painting of Oxford, England


  1. “since telling students that grades and performance don’t really matter would be disingenuous and unconvincing,”
    Why would it be disingenuous? The average job today absolutely does not require a college degree, any more than it did a century ago, when few people went to college except to get training in one of a few professions. GPA doesn’t “really” matter, in any objective sense, and certainly after your first job no one ever cares. Why would it be so terrible to admit that several decades of “if you don’t go to college you’re a loser, and if you don’t go to an ‘elite’ college you’re a loser, etc.” has been catastrophic and has led to all the “fear” you point to? If you’re stuck for four years from 18-22 taking classes that you know in most cases don’t matter at all, in order to become an “adult” in our society, why wouldn’t you try to minimize the effort required to get that piece of paper? Seems pretty rational behavior to me.
    (note: I went to an “elite” college and have done very well, I thrived in the system, am horrified by the pressures my kids have felt that didn’t exist a few decades ago, and appalled by the damage done by people who aren’t purely academic oriented.)

    • Mostly agreed, Brian–and I say “mostly” since I’d argue that the “who should go to college?” question is a related but distinct question, from the “how should we act once we’re there?” question this essay addresses. It is interesting, too, that my experience teaching college history differs in a significant way from the author’s. My students are mostly too indifferent to bother with cheating, though it’s possible that ChatGPT might change that. I’d agree that my students have fear as an important motivator, but their response to that is largely of the passive variety. They just don’t expect anything to make sense, and so they assist in creating the conditions in which that assessment is plausible.

  2. I forgot who said “Never assume malice, when incompetence might be a better explanation.” This essay is a bit overly moralistic, not because it condemns cheating but because Dixie seems to ignore alternative factors.

    Ask yourself: How original is a high school essay written by an honest student? Much of it is basically regurgitation and rewording anyway. What would a teacher expect? What actually PROVES that a student has obtained/retained knowledge?

    I read a recent quip from a neuroscientist about ChatGPT who mused that it seems successful or “hard to discern” because most of our thinking is habitual: we have become like robots (or parrots echoing tweets) so a program that mimics such superficial thinking suddenly seems like a threat (or a breakthrough, if one is a tech evangelist).

Comments are closed.

Exit mobile version