Skip to main content

Verified by Psychology Today

Infidelity

Will ChatGPT Exacerbate Cheating on College Writing?

Probably less than you think.

Key points

  • There is a small constant of students who cheat in college and ChatGPT is unlikely to change that.
  • ChatGPT can help students brainstorm writing topics and get a Wikipedia-depth knowlege on it.
  • The new AI is unlikely to threaten the human creativity required to compose good writing.
  • Educators will likely have to embrace, rather than resist, this new technology.
Hariadhi; wikicommons images
Source: Hariadhi; wikicommons images

College educators are understandably concerned that the new artificial intelligence, ChatGPT, will exacerbate the already serious problem of cheating in college. The AI, launched in November 2022, is the most advanced chatbot to date, capable of understanding natural human language and engaging in (almost) humanlike conversation. The fear in academia stems from the realistic possibility that college students will use the AI to write their essays for them, making the age-old required freshman writing seminar an empty exercise and rendering college writing instructors (like myself) obsolete. Gulp.

ChatGPT is less threatening than you think.

But these fears might be overblown (she says, hopefully). It’s likely this new, exciting technology will enjoy an acute spike in interest and media coverage, only to die down quickly as the next Big Thing in tech comes along. ChatGPT is fun to play around with (if you haven’t tried it out, it’s free and easy to use once you create an account), delivering quick, Wikipedia-like explanations on things; but, at least in my limited interaction with it, it feels more chatbot than human, repeating some canned responses several times in the same short “conversation” and I wasn’t tempted to ask it to write this post for me.

To test out ChatGPT, I asked it to write a paragraph on the subject of white privilege, which it did. Then I asked it to rewrite that same paragraph with relevant in-text citations, which it did in seconds. Then it created a list of APA-style references of those citations at my request.

So, will students use ChatGPT to write their essays? On this question, John Warner, a former university teacher who's authored two books on writing, claims that, “If you create an atmosphere where students are invested in learning, they are not going to reach for a workaround. They are not going to plagiarize. […] But the work has to be worth doing on some level.”

But students who believe the work is worth doing aren’t the ones who cheat, and good teachers always believe the work they assign is “worth doing.” My guess is that the same students who would copy from an online article on white privilege are the ones who would turn in ChatGPT’s “thoughts” on the subject as their own. It’s just one more tool in the cheater’s toolbox. But my guess is that if students have convictions about white privilege, they’d rather write them down than have an AI speak for them.

Plus, new safeguards will be developed. Already, a Princeton student got to work after ChatGPT’s launch designing a program that can, he claims, distinguish between AI-crafted and human-written essays. The coevolution is already underway. It’s interesting that we’re already wrapped up in questions concerning use and abuse of this new technology when I can’t even really understand why it was created in the first place.

Is ChatGPT just a new tool for old-fashioned cheating?

The type of courses I teach, philosophy and writing, are writing-intensive (e.g., several short papers, a few longer essays, some produced in class, most outside of class) and I tell my students that their writing is like their fingerprint: I get familiar with it very early on in the course, homing in on a few problem areas that can be improved over the course of the semester, and I look for steady improvement. Given this premise, cheating in writing is fairly easy to spot. It’s a sentence, paragraph, or entire essay that doesn’t match their writing “fingerprint.”

Many years ago, I had a mediocre student who was not interested in philosophy and not a good writer turn in an essay with brilliant insights into the Enlightenment philosopher Immanuel Kant. I got curious. I lifted a few particularly suspect sentences from her essay and plugged them into the newish search engine, Google, and voilà! Someone’s doctoral dissertation on Kant came up. She’d kidnapped several passages from it and pasted it into her Introduction to Philosophy essay.

More recently, in my freshman writing seminars, I've identified plagiarism several times from sentences that begin with “We discuss” and “We conclude that” when the author is a single student. Here, the student is robotically copying directly from a published research paper written by multiple authors. This move demonstrates both disinterest in learning and astounding laziness; simply changing the pronoun from “we” to “I” might be enough to slip past less discerning eyes.

In all of these instances, in case you’re curious, the student fails the assignment (and sometimes the course, depending on the severity of the infraction) and is reported to oversight committees at the university, which then track them and their work for the remainder of their time at the university. I always take these cases seriously because, while it’s not a crime to be devoid of original thought, it is wrong to try and pass off others’ thoughts as your own. It’s certainly anathema to philosophy, where the whole point is original thought.

Embracing, rather than running away from, ChatGPT.

Technology has evolved so much over the past 20 years I’ve been teaching college, making it easier to copy material from the internet and, in my experience, there’s always been a small but steady percentage of college students who cheat, no matter the course, assignment, time allotted, etc. Simply put: most students just do their work, and a few just don’t, and I don’t believe any new technology is likely to shift this balance much, if at all.

One way to get out ahead of the issue, as educators, is to invite students to interact with ChatGPT in class, researching different topics and assessing how effectively it mimics natural human-like responses. Students can be taught how to use the AI to generate references and citations, and then how to find those original sources through further database searches.

In other words, we’ll likely have to embrace rather than run away from this new technology and find ways for it to enhance what we’re doing in the classroom with students. Some students will inevitably use it to cheat on their writing assignments. But ChatGPT has its own writing “fingerprint” (one I might call “chatbot speak”), which should be easy enough to distinguish from imperfect but original human writing.

References

Cerullo, M. (2023, January 6). Princeton Student Says His New App Helps Teachers Find ChatGPT Cheats. CBS News Money Watch. https://www.cbsnews.com/news/chatgpt-princeton-student-gptzero-app-edwa…

McMurtrie, B. (2022, December 13). AI and the Future of Undergraduate Writing. Teaching experts are concerned, but not for the reasons you think. The Chronicle of Higher Education. https://www.chronicle.com/article/ai-and-the-future-of-undergraduate-wr…

advertisement
More from Liz Stillwaggon Swan PhD
More from Psychology Today