Yves here. The finding in this article, that a lot of young people today will ‘fess up to cheating, is disturbing on multiple levels. First, it suggests that many if not most regard school as an exercise in credential, as opposed to skill, acquisition. Second, it creates a criminogenic environment, in that if a student does not cheat, he is competitively disadvantaged. Third, it points to a wide-spread decline in ethics. No wonder the US has dropped in corruption perception rankings. It looks to be well warranted.
By Austin Sarat, William Nelson Cromwell Professor of Jurisprudence and Political Science, Amherst College. Originally published at The Conversation
My colleagues and I recently spoke with a group of talented, interesting students who just completed their first year of college about using artificial intelligence as a research tool.
I asked what must have seemed like an unrelated question: “How many of you cheated in high school?”
Most of the students raised their hands. Perhaps comforted by the realization that they had plenty of company, they seemed neither embarrassed nor ashamed.
This is not the first time I’ve asked my students that question. On each occasion, the results have been pretty much the same.
By the time students end up in college classrooms, many have encountered cheating and think it makes sense in some cases to do so, because of factors like pressure to succeed.
Let’s be clear: AI has not created the problem of intellectual dishonesty among this generation of students.
Alas, the problem long predates AI and runs much deeper.
The Cheating Pipeline
Many college students are honest and hardworking. But by the time some students get to college, they have become accustomed to academic misconduct in American high schools.
As Eric Anderman, a scholar of educational psychology, wrote in 2018: “Academic cheating is prevalent throughout all types of American high schools. Data from one large national study indicated that 51% of high school students admit that they have cheated during a test.”
Other research on high school cheating found in 2020 that 64% of 70,000 high school students across the country admitted to cheating on a test, and 58% admitted to plagiarism. Approximately 95% of high school students, meanwhile, said they “participated in some form of cheating, whether it was on a test, plagiarism or copying homework.”
And in one Pennsylvania high school, 90 of the 100 respondents to a 2018 school survey “admitted to cheating on some form of schoolwork at least once.”
One of the respondents put it simply: “Everybody cheats.”
Students can cheat for different reasons.
They might feel unprepared for an exam or paper, but they still want to get good grades and gain admission into a competitive college.
They might recognize that cheating is wrong, but they justify it by saying everyone else is doing the same thing, or that they have teachers who don’t do their jobs well. Other students might not fully understand what cheating means in different contexts or think that what they are doing counts as cheating.
This kind of thinking can allow students who sometimes cheat to not think of themselves as cheaters.
Sociologists Gresham Sykes and David Matza call this tendency “techniques of neutralization.” This means people use their internalized ways of seeing the world to justify acting in a way they know is wrong.
Looking the Other Way
A 2020 study of 840 undergraduate college students found that 32% of them had cheated in some way on an exam.
College professors like me may be tempted to look the other way if we suspect a student is cheating, or try to solve the cheating problem by changing the ways we evaluate students.
The Wall Street Journal, for example, reported in 2025 that faculty across the country are giving up on writing assignments, which students can produce with AI, and returning to in-class tests and examinations.
Every college and university has rules against plagiarism and other forms of intellectual dishonesty.
To offer one example, Harvard’s policy says that “Cheating on exams or problem sets, plagiarizing or misrepresenting the ideas or language of someone else as one’s own, falsifying data, or any other instance of academic dishonesty violates the standards of our community, as well as the standards of the wider world of learning and affairs.”
Students who violate the cheating rules at Harvard and elsewhere might face consequences ranging from failing a class to being expelled. But many instructors don’t report incidents of cheating to administrators responsible for enforcing those rules and meting out punishments.
Few colleges have developed an intellectual integrity curriculum that treats cheating as a habit and works to counter it over the four years of a student’s college education.
I think that, like any bad habit, students can only be weaned from cheating slowly, with a support program and clear, severe consequences when they are caught.
Cheating in College
Getting a sense of the dimension of the cheating problem on college campuses is not hard.
In February 2026, for example, a Harvard undergraduate student named Matthew Tobin published an opinion piece in the Harvard Crimson entitled “Plagiarize or Perish.”
He cited a 2024 Harvard Crimson study that showed 47% of 850 surveyed senior students said they had cheated.
Tobin wrote that while some people say cheating is the result of “modern students’ scholastic disengagement or use of artificial intelligence,” other issues are at play. Plagiarism and academic misconduct “have been happening all too often at Harvard for far longer than the advent of these issues,” he wrote.
Reported academic misconduct cases increased at Ohio State University by 57% between 2014 and 2018. This is likely a low estimate, since most academic misconduct cases are not reported or investigated.
Charlie McLaughlin, an Oberlin student, published an op-ed in the student newspaper in May 2026 criticizing the college’s decision to change its honor code charter to allow professors to proctor tests, meaning supervise students while they take the exam.
“Changing this policy is a clear sign that this school doesn’t trust us to learn to be adults with integrity,” McLaughlin wrote. “That’s sad. Maybe, it’s also reasonable. Maybe, we don’t deserve that trust. That’s even sadder.”
Princeton also recently abandoned its 133-year-old prohibition against proctoring exams “to address increasing concerns over academic integrity violations, including the proliferation of AI usage.”
A Teacher’s Dilemma
I don’t think of my students as cheaters, and I don’t want to regard them with the kind of suspicion that turns teaching into a policing activity. But it is my job and that of the college where I teach to recognize that our students need a lot of help to develop good academic habits.
Unless colleges acknowledge these facts, I believe they have little chance to curb the pervasiveness of cheating.
Faculty can start by weaving discussions of intellectual integrity throughout their courses and enlisting students to think about who they want to be – and whether they want to live their lives cutting corners and gaming the system. Only then can colleges hope to build what Tobin calls “a commitment to academic integrity in (our) students.”
Skating through elite institutions without actually learning much is precisely how we wound up with CEOs bringing in “AI” to their companies without having any valid use case for it ahead of time, to name but one of the legion of sins of the PMC class.
I’m reminded of Stoller noting in his book “Goliath” that a very large cohort of the New Dealers from the FDR era who turned the country around for the better came from state schools, however it sounds like the rot has reached there as well.
Wright Patman was a star of Goliath. Naturally, the first class of Tech Bros (post-Watergate Atari Democrat dumbasses) pushed him aside. Patman went straight from high school in Texas to law school at Cumberland University in Lebanon, Tennessee. Those who became the “Best and the Brightest” didn’t impress him, much.
Define cheating.
Is writing an in-class paper without having read the book cheating?
Is looking over at someone’s paper once or twice the same level of cheating as going into an exam with a prepared set of answers?
Whatabout knowing someone will look at your paper, and not doing anything about it out of sympathy or intimidation?
There is a difference in kind, I think, between pulling one desperate cheap move and an established campaign with time and effort on the
groundwork for deceit.
The Harvard Study has a 30% number for using AI when not instructed to do so. Probably mean for the bulk of a paper of while answers, but it’s not clear.
Everything you’ve written here is an example of the “Techniques of Neutralization” that was linked to in the text of the article.
Cheating is anything that is considered a violation of the academic integrity policy of the course or the institution. In the same way that even though speeding and drunk driving are crimes that fall under the umbrella of “reckless driving,” with differing severity and different sanctions, the purpose of traffic laws and using the criminal justice system to enforce them is to prevent both kinds of reckless driving from happening, and identify culprits and remediate it if it does happen.
Similarly, even if “looking at someone else’s test for their answers,” is less egregious than “sneaking in your own prepared answer sheet,” they’re both bad, should be prevented, and if one is caught doing either, result in sanctions and interventions, albeit with less harsh punishments or a longer sanction escalation for sneaking peeks compared to wholesale plagarism.
I went through my undergraduate education at a time that smartphones were starting to become widely adopted. I got a Blackberry as a gift for my 21st birthday, and others were starting to use iPhones. Our exams for multiple classes were proctored, and we generally had a three strikes rule for those with wandering eyes. First time was a warning to not look at one’s phone or elsewhere if still submitting questions. Second time was a request to either turn-in the exam as is, or relocate in the exam room to help isolate the student and ask them to surrender their device to the proctor. Third time was a forced end to taking the exam and referral to the department’s dean for academic integrity issues. Comparatively, more egregious cheating like bringing in a pre-written answer sheet resulted in test confiscation and referral to the dean immediately.
Cheating is a societal norm in the USA, and has been for some decades.
I was once kindly informed that I “Lacked the ethical flexibility to succeed in this business”.
It was meant in a kindly way, and it was true.
The advantage goes to the individual who successfully cheats, the costs are borne by Society as a whole and those costs are substantial.
I think you should be proud to lack ethical flexibility
Yes this is a societal issue. The ends justify the means for a lot of people, and allowing teenagers to follow along with their example is like asking a Democrat to cave. Yes, just give them an excuse to.
You can’t have the kind of leadership we have and expect anything else. The fish rots from the head!
The education-to-employment road has also changed. In 1965, a UK electronic engineering graduate would be hired by a company like Plessey or Rcal, and expect to be there for their whole career, working their way up. And some part of the first year as a raw graduate would be spent in different departments for a week each to understand the company. So… bad education, lousy prospects.
Now, PMC engineering employment provides no job security. No respect by the institutions leads to no respect for them. Blag one’s way into a job, perhaps sink or swim, and see how it goes. If necessary, move on.
Cheating in finish carpentry: “Putty and paint make it what it ain’t”. However, it will become what it is after a short time in service, and you will get a call-back. Then, if you get a reputation for sloppy work, you will not get work, especially with on-line comments. Probably not applicable to college+ cheaters. Seems, the more they cheat, the higher they go.
Glad you brought up poor workmanship in trades. We did a lot of the construction work on our home 47 years ago (much stronger and less tired then!). So we know what quality workmanship looks like. Back in 1979 most all of the work we hired out was done well (the alcoholic electrician being the sole exception). In the past few years we had to replace windows or roofing, did updates, replace appliances or mechanicals, etc. What a difference a few decades made. Cutting corners seems endemic now and quality of manufactured items seems mostly worse. I’m not sure posting bad experiences on social media roots out “cheaters” in building trades. There is a very real shortage in trades and some of these guys apparently care less about getting repeat business; maybe general contractors take note, however.
Micron is doing site work on a chip fab locally… we’ll see how that goes. The state is rebuilding I81 through the region and it is not uncommon to see concrete work being ripped out because of noncompliance. These major projects are typically years late and over budget. Cheating and incompetence plays a big role.
I was a graduate assistant at a state school about a decade ago while working on my degree. I did take action on a couple cheaters during my time but I don’t think the practice was widespread. Easier to turn a blind eye for most professors/teachers.
Another part of the issue, not mentioned in this article, is that schools do not support professors/teachers enforcing the penalties for cheating (i.e. failing them, expelling them, etc.). My university treated the students and, I quote, as “customers” so there is a strong incentive to not punish academic dishonesty.
As Yves said, this is definitely NOT good for society at large.
It’s a problem in Europe as well.
Ultimately it comes down to incentives and ease of accomplishment. I don’t think students are different from any other cross-section of society, in that given an easy way or a hard way, they will generally choose the easy way. The fact that some may choose to justify themselves by arguments about the nature of society, education and so forth are essentially just updated versions of “the dog ate my homework.”
Accepting that many students will cheat if they can, the question is how easy or how difficult you make it for them. Banning smartphones from written exams was one simple way of doing this. Anti-plagiarism software is in my experience now so effective that it acts as a real disincentive to cheat on written assignments, though there are still those who think they can get away with it.
In spite of what the authors suggest, AI is fundamentally different in that it brings together the maximum of temptation, the maximum of opportunity and the maximum economy of time and effort. It’s also much more difficult to detect reliably, and, unlike plagiarism, it’s almost impossible to prove beyond doubt. It’s also much easier to defend (AI did the first draft, I used AI to find some references, I asked AI if the text could be improved) and which are again effectively impossible to disprove.
But I think the battle is being lost, unless we go back to the pen and paper days of my youth. The problem is that like all superficially easy and convenient solutions, AI has massive drawbacks. We’re now looking at a generation exiting university which has never really learned how to do research, nor write structured and organised prose. And if the AI boom goes bust, as it looks as if it will, they really will be in it.
This is nothing new. When I was in college in 1969, I remember being in a frat house where someone showed me a file cabinet filled with term papers that were graded well, all from previous students who had graduated over the years. If a current student at the fraternity needed to submit a paper, he could easily go through the file cabinet, find one that fits, copy it on a typewriter and hand it in.
Not that long ago the advent of online learning platforms (like Mastering Physics or EdX ) were a welcome innovation to help deliver homework and online exams/quizzes for huge first year classes.
Then sites like Course Hero or Chegg arose that bought and sold answers to questions scraped from the web. Students would simply look up the answers in one tab with the HW or quiz open in the other. Now with Artificial Information, they just copy and paste the question and get their answer ( not always right btw). Can’t tell you how many times a student has come to me with help on an assigned question THAT WAS MARKED CORRECT. They had no idea how to do it.
Thats IF they come to office hours. They’ve stopped coming and told us they just use AI.
Because so many students are getting near 100% on their online HW and quizzes, but 50% on their handwritten exams, we’ve abandoned online assessments and use mostly in class ones for the bulk of the grade. (Retaining some nominal grades for online work to get them to do it. Not everyone cheats, and some retention for those that do is better than none)
Pencil and paper with no devices as they sit in front of you is the only way. What was old is new again.
I taught a college STEM class for the last 6 years. My class could not use AI since it was making engineering drawings so they had to learn 2D or 3D CAD so I didn’t have to deal with the cheating, but I did have to deal with “disabled” students. Which is where I saw a problem.
They handed out these “disabled” exemptions like candy. We would be informed we had so many disabled students in our class. We would contact them as ask what they needed to succeed in the class. Most didn’t even answer. We had to give them extra time, tutor, or whatever they wanted to make sure they made it through the class. Whatever it took.
Some had legitimate handicaps such as hearing, or a physical handicap where using a computer made it difficult. No issues with any of that. But, many, and I would go as far as saying most, didn’t need anything. They were perfectly capable of doing the work like everyone else, but they wanted the “disability” exemption to skip class, be late on assignments, and do as little as absolutely possible. And we were expected to pass them. Sports people were pretty much the same.
I ended up in the Dean’s office due to one kid who missed 11 of the first 20 classes and 7 out of the last 8 and somehow it was my fault. I didn’t have enough empathy for disabled people. I told the Dean there was nothing wrong with this kid. He was nothing but an excuse factory playing us all for chumps getting the disabled tag. That didn’t go over so well. I didn’t care and was part of why I left. I can’t teach people who don’t care about learning. Unfortunately, life will sort these kind out.