The Integrity of Academic Integrity Enforcement
The philosophy professor heading up the office at his college that handles student violations of academic integrity, like cheating, is dealing with a seemingly stonewalling upper administration, and needs some help.
Further, there’s some reason to think that political considerations are playing a role in the administration’s inaction.
The professor writes:
I’m in need of advice from the readers of Daily Nous!
I’m a philosopher who serves as an academic integrity officer at a public liberal arts college—and I have a problem. Several faculty members have reported to me that they’ve discovered two students who have been systematically cheating in their classes by generating their papers using AI, when this is expressly forbidden. (One of the students did this in several philosophy classes.) Some of these classes were from last semester, but our policies explicitly allow students to be charged for past infractions if new evidence comes up, as it did in these cases.
So, no problem, right? The academic integrity cases get filed, and things take their usual course.
Except…. not this time. The administration is simply refusing to have these cases heard. They’re also simply refusing to say why they’re not having these cases heard. And, by the way, there is no possible doubt that the students are in violation.
I suspect that what’s happening here is that the administration is caving to parental and/or political pressure to give these students a free pass. The parents of one of the students contacted the Governor’s Office, who then sent at least one email to the administration on their behalf. I’m not privy to what might have happened here. But it looks like at least one of the students is getting a free pass on multiple counts of cheating because the administration is worried about parental/political response if they’re subjected to the usual academic integrity proceedings—proceedings that would likely preclude them from graduating this May. I suspect that the other student is benefiting from this. The cases are very similar, and so if one student gets a pass then the other one does also.
I’m loathe to just go along with this, toe the administration line, and ignore the students’ infractions. Students shouldn’t be getting college credit for philosophy classes that they didn’t earn.
TL; DR: Two students have clearly cheated in several classes (including several philosophy classes) and the administration is refusing to have their cases heard. What to do?
If you have any advice for “Professor Double Trouble” or experience with these kinds of difficulties, please let us know about it. Thanks.
I don’t have direct experience with this type of case, but my first thought is that Prof DT should very carefully review all applicable policies related to confidentiality, etc., so that his own actions are legally and institutionally unimpeachable.
Sadly, in this instance, there seems to be more professional (and personal) risk confronting the failure to follow institutional policy than in letting it go (public institution and all that). But, there are possible remedies through the accreditation process (not following institutional rules)–not ideal nor expeditious but ultimately some pressure could be brought to bear. Slow perhaps, but less fraught.
Shame to have to hide the institution that is doing this! Understandable that everyone’s looking out for their own hides and institutions nowadays, something about contracts and ethical standard… but moments of fundamental conflict like this are only increasing in frequency and the fact that people experiencing them are forced into the shadows of a philosophy forum to seek advice really isn’t all that encouraging. As others have said it’s essential you’re protecting yourself (assuming that’s what you want most in the end and don’t mind not being able to discipline students using AI in the future and that doesn’t make you feel awful about the integrity of the institution and the point of your effort and the efforts of your colleagues), so it seems like the best thing to do is nothing! Especially if the alternative is becoming a liability to the university somehow at the expense of feeling more principled or “doing the right thing” or something. These political conflicts will only be more common in the future (especially as federal funding & accreditation processes are hollowed out, too)… a bleak question indeed. No good answers. Best thing to do may have been just this, submitting anonymously online alongside (hopefully) many other administrators and professors experiencing this type of systemic disintegration, making these events known. Thank you for speaking out.
I’d imagine several newspapers or news magazines might be interested in doing the legwork and following up on the story / shedding some light on it, perhaps as part of a larger story dedicated to AI and cheating. This is especially true if the Governor’s office has sent an email to put pressure as was indicated in the post (moreover, since this is a public institution, FOIA is a great if somewhat slow tool if one knows what “search terms” to ask for if / when the paper requests any “responsive documents”). There are enough data points here (e.g. public liberal arts, May graduation date, philosopher working in the academic integrity position, etc.) that a decent investigative journalist shouldn’t need all that much more information in the tip (e.g. the state it is in) to be able to reach out and get the ball rolling on sheding some light on the situation. The NYT has a great “tips” procedure online, as do several other papers like the WaPo or whatever paper is regionally going to care the most. I’d recommend getting in touch with them (not from a work email) or submitting an anonymous tip so as to get the ball rolling. Integrity is important. So is finding ways to bring powerful actors to account when they are behaving badly and using illegitimate political pressure. It may seem extreme to contact the press, but if the parents have contacted the Governor’s office then that is already quite extreme.
I don’t know how faculty governance is structured at Professor DT’s institution. But at my institution, the smart move would be to bring the matter to the Faculty Senate–who has ultimate authority over academic standards.
Even if the Faculty Senate does not have authority over academic standards at Professor DT’s institution, getting them involved would at least make it the case that upper admin can’t just let the cheaters off the hook, bury the case, and quietly move on.
Agreed. Also, if there is a chapter of AAUP/AFT, check with them. Be careful not to violate FERPA. Check your faculty manual to see if decisions related to academic integrity are in the hands of the faculty.
This is going to sound like a joke suggestion, but it’s not.
I think, for all of your own protection, faculty across different campuses should publicly commit to all using an LLM – could be ChatGPT, could be some other – to grade ALL student’s work. Let the students know, in advance, that this has now become the policy for every department, on every campus. If students can use AI to lessen their workload, then it is only fair that faculty be allowed to do so also.
Should there be NO outcry, then great, faculty can begin using ChatGPT to assign random and meaningless grades to work which is increasingly becoming random and meaningless. I think that’s a fair trade, as it would allow faculty to have more time to focus on teaching students, doing research, etc.
More likely, should there be an outcry, driven primarily by those students who care, and who actually do their own work, and made a political matter by those student’s parents, then this outrage can be leveraged by faculty to ask the following pointed questions.
Why should faculty be prevented from using AI to do their work, but not students? As in, why can’t faculty use AI to mark the work of their students? Or to determine what the academic fate of a student ought to be, given that they have been caught cheating?
If, in any way, the public’s response is something along the lines of – we cannot trust AI to do as good a job as humans, then faculty can respond by noting that that same logic applies to the use of AI to do the student’s work as well. And now that we all agree that AI cannot reliably be used to do the student’s work, and given that we already all agree that it’s not fair to give anyone a free pass, then let’s work together to put into place policies which will actually enable faculty to hold students accountable.
To put all this another way, if we can get all of the OTHER students and parents upset about this issue, such that they too are ALL calling the governor’s office, and putting pressure on the administration, then it puts not only those folks in a difficult position, but also those politically connected folks who see no problem with trying to grease the wheels for the sake of their cheating children.
College administrators live in fear of legal liability. You need to inform them that their actions are creating a plaintiff class for a class-action liability suit against the College. The plaintiff class is constituted by those students who have been successfully prosecuted for Integrity violations. The College has discriminated against them by prosecuting them when it does not prosecute other students who are equally liable to prosecution.
Furthermore, you as the head of the office are open to litigation. Talk to your Corporate Counsel’s office. Tell them that the Administration’s actions are making both the College and yourself liable to suit. Explain that you want full indemnification for the actions that the Administration is asking you to take. Explain that you want a written document that will hold up in court, spelling out in full that your decision to prosecute Students A and B was overridden by administrative fiat.
You are probably not going to get the written document. And you may get fired from your job directing the Integrity office. But when Counsel’s office talks to the Administration, they will make it clear that the Administration is on shaky ground. And it may give President Peabody the excuse to tell the Powerful Parents that his corporate counsel’s office has told him that he cannot veto the prosecutions without incurring legal liability.
I have no relevant expertise but I did spend.many decades in academia mostly at a large, state R1 university, about a third of that time doing administrative work, and over a decade doing research on First Amendment freedom of expression issues. Here are some questions it might be prudent to explore:
Do you work at a college where your are tenured and tenure provides meaningful protection?
Does the state in which the college is located have statutes that protect whistleblowers? Answering this question will probably require consultation with an attorney who does not work for your college or for the state. Is there a faculty ombudsman at your college? Are faculty unionized (with a union attorney)?
The attorneys who work for your college represent the college, so I would not consult them first (if at all); they might regard you as a threat to their client, whatever your intentions.
When, during the mid-1990s, I asked Professor William Van Alstyne whether some uses of sexually explicit material for college teaching were protected under the First Amendment, he replied (with appropriate qualifications) that they were indeed protected and that the instructor would eventually prevail after at least several years of grueling court battles and the instructor would then be awarded back pay.. Surviving until then would be the main issue, he said.
What would be the consequences for you if you were to resign from your role as an academic integrity officer? Would you safely be able to tell colleagues why you resigned (assuming of course that the students involved were not directly or indirectly identified by you)? Perhaps that is also an issue best discussed with an attorney.
To find an attorney with whom to consult, you might inquire of a local ACLU office or contact FIRE or perhaps the Academic Freedom Alliance.
You did not say what evidence you have that AI was used.
At my institution, we’ve been warned that colleges are losing lawsuits when the only evidence they can present documenting the use of AI is TurnItIn or GPT0 or some other service’s assessment. We need “smoking gun” evidence to level an accusation that may obstruct a student’s advancement. Employers want students to have AI-skills, so we have no right to object to their acquisition of those skills.
If admin is not backing the faculty in this case, it may mean that a suit has already been threatened, and the relevant question is not whether the evidence can convince a faculty panel, but whether it can convince the court.
The college essay is dead as an assessment of learning.
We should also be moderately skeptical of our ability to reliably detect unauthorized AI use.
This is, sadly, correct. In all 200- and 300- level classes I now use exclusively blue-book exams.
There were clear smoking guns in every essay in each case, including multiple fabricated references and non-existent sources.
Fabricated quotes/references are so far the only true smoking gun I can think of. What did the students say in their defense?
Nothing… as the administration is refusing even to have hearings!
But typically an instructor would meet with a student they suspect of cheating. Did it escalate before they could even talk with their professor?
Then that’s your ground for the integrity action. No need to speculate about how the falsified research was produced.
Out of curiosity, why can’t we just aim for the second best? Like, given that hoping for a real misconduct penalty is hopeless, can’t we just give a failing mark? If the detection of AI is due to BS quality, this warrants a failing mark for independent reasons anyway.
It might be second best, but it’s a distant second. Someone who cheats at university has no business being there, and poses the constant risk of further attempts to get a fraudulent credential. Much worse still: if it becomes known that the university treats clear cases of dishonesty as nothing worse than a failure to learn the material, there will be a much greater overall incentive to cheat, and in fact the honest students may soon find themselves in the minority — and looking worse than the cheating majority because of their integrity.
Granted, but this is exactly where I feel the hopelessness. In my institution, students already know the necessity to check whether references are fabricated. And since we can no longer use this indication as decent evidence, students have learned to BS and deny any allegations of AI usage during misconduct meetings. For reasons mentioned in other comments, we no longer have solid proof. So that’s why I think giving out a clear failure is the best I can do.
i don’t know if it’s possible given the state of things where you are, but if your institution permits low-friction retroactive grade changes by instructors (rare i think), you could try going back to correct grades (or having the instructors of record go back etc.) you think were fraudulently obtained.
in my experience deans and registrars are very touchy about the responsibility for the final course grade so perhaps a minimal but real form of leverage you have is to disown the fraud and make them take responsibility for it.
perhaps they are more willing to let the processes work than they are to arbitrarily go in and ‘de-correct’ grade corrections?
practical advice:
1) make secondary sources optional when they aren’t necessary for your learning goal.
2) if you do require them, make sure that in-text page number citations are required. present llm hallucinations fabricate the material in the real source and assign random page numbers, so double check the pages cited when possible.
quote from assoc. dean on recent finding: “The instructor provided an extremely detailed description of the allegation. Most helpful in terms of ourpreponderance of evidence standards was the mention [not mention, documentation] of ‘hallucinations’ or false citations.”
Suppose you’re working at the mint, and you notice that management is doing nothing as people wander in to use the dies to make up batches of counterfeit currency. When you say something about it, they all look the other way or come up with excuses. You see that this will devalue all the properly-minted currency, and that it will lead to serious problems down the road. But nobody at the top levels seems at all worried. They are already corrupted and in fact don’t even do much to hide the fact that they are doing nothing to safeguard the integrity of the currency. If they can bring in money this way, they are happy.
What should you do?
It’s a difficult question. A simpler question is: is this organization terribly corrupt? The answer, of course, is yes. Either steps will be taken to correct the problem before the system collapses, or it will have to come crashing down when the rotten leadership weakens its core structure. It ought to be the case that the corrupt people and practices are removed or corrected in time, but if enough people turn a blind eye, or if the whistleblowers are allowed to suffer for their integrity, then there might be nothing an individual instructor can do to bring about a better result.
In the case of the universities, which mint academic degrees, the problem can be more difficult to solve since the debasing of credentials can be dressed up in the costume of virtue.
There should be strong obligations on instructors to ensure integrity. But in a corrupted system, where instructors with moral integrity are under threat for their pains while their corrupted peers enjoy the approval of the higher-ups, it might lead to a better outcome to work for an end to the systemic problems.
Such is the lot of any moral person working in an institution that doesn’t care.
I say let them graduate and be done with it. Who cares? What are they going to do. Their parents got them out of it, Ok. They graduated and cheated in one or some classes. Who cares?
You have the proof, everyone knows, the parents, too. The parents somehow know or have access for these types of ‘favors’, Ok. Everybody else doesn’t.
You really want to make this a big thing? You can. You have the proof!
Socartes died over questions people still debate today. Galileo had proof. There’s no point of dying on this hill. You see tomorrow he’s probably thinking. Who cares?
The question is should his buddy get away with it too?
“there is no possible doubt that the students are in violation.”
“Except…. not this time.”
This is a horrifyingly bad take. Your ignominy is in the mail.
You said, “Except… this time”.
You also said, “no possible doubt that the students are in violation.”
So you have the proof, you can prove it to the parents, the governor, and everyone else.
The parents got their kid out of it, they asked someone for a ‘favor’, and got it.
Socartes died over values and principles we still explore and discuss today.
Galileo’s recant was over facts he can prove. Why die on this hill, when tomorrow you will see.
You can fight this if you want. And win.
What are you fighting here? The fact this student cheated? You have the proof!
Are you fighting this ‘system’ of pulling strings to get a student, a Philosophy student, not an Engineering or Premed student out of cheating?
This thing that parents (who may be disappointed) got them out of it, is not forever.
Is the student going to continue writing papers that can be proven fraudulent and get away with it?
The real question is should you let them both go and just move on?
Ps.
Other students don’t have this privilege, and get punished for it. Maybe that’s something everyone should hear at the start of the semester.
This is obviously the same commenter as above under a different name.
Yes I go by both names. I changed the name Emails.
This wa a clearer response than the last. I posted it right after, to clarify the point.
I’m a Philosophy Student, and will be forever. I question if I love Philosophy, the subject, or am I a philosopher, lover of Sophia.
I also don’t think it’s some
Medical or Engineering Degree, or Pilot Test where of course fighting back is the correct response.
Cheating on a philosophy paper to graduate. What could he possibly do with it?
*** Maybe the moral of the story could be ‘Unless you have connections to get you out of it when you get caught.. Don’t cheat, we will catch you.
To Pageturner.
Am I saying something alien here? Do you exist in the world?
Ignominy and I run into eachother alot actually .
🙂
If the person heads up the committee that investigates all accusations of academic integrity violations, it could be possible to hold up all such investigations until these are investigated.
Although it is not an unsubstantial burden on us, I think that the best way to avoid these issues going forward is to be intentional about designing assignments that are as AI-cheating-proof as possible. Because I am hesitant to abandon the take-home essay assignment for blue book exams, I began employing a technique about which I learned from this blog, I believe: I require students to do all writing work (from outlining, drafting, to final product) in a Google doc that is shared with me (and any TA), as well as uploaded to Canvas. My TA and I are then able to use a free app/browser add-on called Revision History (https://www.revisionhistory.com/) that enables one to replay the composition in real time (or speeded up to 8x) — in addition to assessing any “flags” for unusual activity that the app automatically detects. (I have found Google’s own revision viewing function insufficient because it merely captures various “time slices” of composition, not the entire process.) I instruct students that, although I do not forbid AI, any use will be considered in determining their final grade — with any “outsourcing” of their own thinking regarded as grounds for failure. This has not prevented students from cutting and pasting a near complete draft of their papers into, say, Grammarly, and then pasting and submitting the Grammarly-improved outout as their final draft — but when the resulting changes are in fact simply grammatical I am OK with that. (Noting this, some such students now type into their document “I am now going to cut and paste into Grammarly for a grammar check” to make crystal clear what they are doing.) In other cases, I have detected cheating (for example, by using something called Wordtune to write or substantively rewrite). In cases where I have confronted a student regarding cheating, they typically confess.
Again, this has added substantially to my workload (I ask my TAs to flag cases, but do not burden them with reviewing entire Revision Histories). However, its advantages include making the ethical use of AI a skill cultivated in the course, serving as a disincentive to cheating, and providing me useful information to convey to students about the writing process. For one example of the latter, I can tell students the least, most, and average amount of time that it took their peers to compose “A” papers. (Something I found extremely informative, as well, in gauging the workload I was requiring of students.)
It remains possible, of course, that a student has AI write their paper outside their Google doc and then types the entire thing into the doc. But this, too, should be evident absent the student taking time-consuming means to make the typing appear as if it were an actual composition process (I assume they cannot, without assistance, simply produce a competent 5-page paper word for word without obvious revisions). Again, not impossible, but also not possible without a student likely having to spend much more time cheating than it would have taken for them to simply produce their own work.
Alas, this technique will work only so long as it is not possible to have an AI agent instructed to compose a paper as if it were a student drafting their own work. So far as I know, we are not there yet, though I would be eager to hear if others know otherwise.
“I think that the best way to avoid these issues going forward is to be intentional about designing assignments that are as AI-cheating-proof as possible.”
I thought you were going to suggest that we should rethink the very nature of our assignments, not just how we police cheating. This system certainly provides an incentive not to cheat (fear of detection) but it fails to address the incentive students have to cheat in the first place, absent such safeguards. If without policing they can’t be bothered to write their own paper then maybe we are due for some self-examination as well. I say that as someone who deeply values the take-home paper but is willing to admit that we may all be looking at this with status quo bias.
I agree. I focused on paper writing because that was the kind of assignment that prompted the original post (and, again, because I am hesitant to abandon the advantages of paper writing). That said, I agree other measures may be necessary to address “the incentive students have to cheat in the first place.” To that end, in all my classes of 40 students or fewer I have transitioned to “Harkness” style teaching. All students are assigned to a discussion group of 8-10 students and, for each class, submit short answers to text-based “Prep Sheet” prompts in advance of class (assessed on a “complete/incomplete” basis). At least half of each subsequent class session is then devoted to them discussing their responses to the prompts in group discussion (with me floating among the groups for guidance and assessment). I find that this exploits the positive effects of peer pressure — they do not seem to want to appear unprepared to discuss with their peers. Neither do they seem inclined to try to pass off AI-generated responses as their own in this setting. I am old enough to never have had in-class discussions of this sort during my own higher education. Now that so much content is available to students outside of class, however, I find providing opportunities for guided in-person discussions of this sort a more productive use of class time than lecturing, for example.
To your last point, it would not take an overly savvy undergraduate student to have an AI-assisted program do this for them. I imagine this is in the minority but it will and can be done quite easily, especially as AI-tools do all the coding for people.
I hope my undergraduates do not read this blog! Thanks for letting me know. I followed up and indeed it seems that Google’s Gemini already has such a feature (by invite to “Workspace Labs”?) in Google Docs. For now, my university does not provide access to Gemini to students, so — if I understand — they cannot (yet) use this feature when logged into their university account for Google Docs. Nonetheless, it seems that the ability for them to hack my assignments is likely to arrive sooner rather than later. I just may retire the take-home paper after all. Sad.
There’s probably nothing that you can do other than have the professors fail them in the respective classes for cheating. But from what I understand most universities don’t allow academic punishment for using AI since they can’t trace it like traditional plagiarism unless the student admits to using AI.