A Tired Topic: ChatGPT and AI in the Writing Classroom

24 July 2025

Rant Warning

I wasn’t planning to drink wine tonight—let’s say my students drove me to it.

As an instructor of College Composition I and II for the past decade-ish, I’ve only rarely ever encountered the occasional case of clear-cut plagiarism. And with plagiarism detectors aplenty at my fingertips, when those essays did cross my desk, there was really no question about it. Moreover, it was the rare scribbler, indeed, who’d have the audacity to try to dupe me, when we stared into each other’s faces two or three days a week.

With its rapid evolution and rise in accessibility and popularity, you’d think ChatCPT would have crossed my (every moment more) virtual desktop far oftener to date than it has. And at this moment, the only thing I can attribute this luck to is my in-person presence and credibility in the classroom. I humor myself, perhaps, but I think my students respect me when they can see my face, know I’m a real person, and understand that I’m here for their benefit.

This is true, too: I do my job for their benefit. The education gods (and everyone else) all know that I do not make enough money at this job for a living. I am lucky enough to have a husband whose job covers most of our bills. I’m in those classrooms for my students—because until today, I have been in love with what I do.

This summer session, because our family is moving and I had residency in Denver, and for a dozen other reasons, I took on two sections of 12-week Comp I online. This is never my preference. I need faces, and I think my students do, too.

And now, in the past two weeks, I have witnessed no less than eight blatant instances of either plagiarism or ChatGPT or some grotesque combination of both.

Look, I’ve done my fair share of therapy, mindfulness work, and Buddhist studies. I recognize that if I allow this—pandemic—under my skin, only I am to blame for the frustrations it evokes in me. I own my emotions and my reactions. I do.

But y’all.

I am mad.

I think?

The first student I caught using ChatGPT might be exempt from this emotion. Here’s what happened.

This student is nonbinary and goes by they/them pronouns. Let’s call them Alex (not their real name). Alex submitted a Writing Workshop for an essay that included MLA citations from outside sources and an MLA Works Cited page. It is always my goal to help students learn their citations, so as always, doing more than I probably should for my job, I went to find Alex’s resources on the web to ensure that they were citing them properly both in their in-text citations and Works Cited page. Now, these were full-on quotations with parenthetical references and correlating Works Cited entries, mind you.

Except they were all bullshit. Two of the Works Cited entries did not even exist out in the real world. The third entry I could locate, but in it I could not find the quote Alex had implemented in their essay.

At first, I thought it was my error. Maybe I needed to check the scholarly databases: EBSCO, Gale, Wiley, etc. Nada. Alex included URLs for all three sources: dead ends, “Item Not Found.” I just sat there, shaking my head. How could there be sources listed with literal quotes and parenthetical references, but I couldn’t find them?

Then it dawned on me: none of it’s real.

None of it was real.

None of it.

I gave the student a zero on the assignment and requested they join me during my virtual office hours (Zoom). To my eternal surprise, they showed up! They showed up, cow-towed, head low to their chest. I barely saw their eyes the entire meeting.

But they admitted it.

“I ran out of time,” Alex said. “I really didn’t want to fall behind, and I had so much going on this week, including a death in the family, and it was just so much faster. This way, I could just get it turned in and move on.”

In response, I reminded Alex to simply ask me for an extension next time.

BUT OH, MY READERS, THIS BROKE ME!!!

What a freaking disaster of a world my poor students live in! They’re being asked to engage 2-6 hours per week in my course, when ChatGPT can produce a “quality” result in 20 seconds. I almost can’t even blame them—can you? Except, I can.

Well, not Alex. Alex has redeemed themself.

Today, I held a meeting with another student I suspect of ChatGPT use. This student, however, never admitted anything. She kept saying, “Yeah, I wasn’t sure I was doing it right,” and “I kept rushing to get it done.” But after some instruction, she admitted she’d never once watched any of the instructional videos on the online course. And after we concluded our instructional meeting, I took a moment to look up her sources: they too were fabricated.

My husband, Brian, and I have had discussions about ChatGPT in the writing classroom quite a few times. Often, students will submit writing that is clearly ChatGPT-produced. They don’t know enough about writing to understand the impossibility of a Composition I student leaping to a PhD student’s level of writing in one week. I’ve been in an MFA program with a brilliant cohort of other writers, and not a single one of them have made that leap in two years! It’s not possible.

But this is a whole new level of distrust for me. Suddenly, ChatGPT is feigning Works Cited pages, resources, quotes, summaries, citations? Don’t even.

My close friends, Jim and Vineeta, keep extolling the virtues of ChatGPT in terms of making our lives easier. And they may have a point—to a degree. But the way I see it is this: when we alleviate all of our struggles, what strength can we develop? I won’t avoid osteoporosis during menopause by lifting light. I won’t improve my cardio health by parking in a handicap spot and refusing to walk, if I am able-bodied. And if I do not use what rattles within this skull, like my sweet Nana, I will lose it.

I did have a moment of weakness myself, on Tuesday. My mentor asked me about a month ago to develop a logline for my novel. With all the chaos in my life now, I put that assignment off, then further off. On Tuesday, I decided to do it—except, my brain was exhausted. So, in a moment of whimsy, I plugged the details of my novel into ChatGPT and asked it to create me a logline.

Here are the prompt and results.

ChatGPT Prompt:

Write a logline for a novel about a young military spouse and new mother who loves literature and has borderline personality disorder and daddy issues and tries to have an affair with a photographer while her husband is deployed 

ChatGPT Logline:

While her husband is deployed overseas, a young military spouse and new mother with a deep love for literature, a fractured self-image, and unresolved daddy issues teeters on the edge of self-destruction as she becomes entangled with a charismatic photographer—forcing her to confront the chaos within before it consumes everything she holds dear.

At first, I was flabbergasted by this. It looked great to me! Like precisely something I’d read on a novel’s jacket. But then I sent it to a friend, and he and I began to pick it apart. Was that two, three, four clichés in a row?

“You can do better,” my friend said. “It’s not that good. It’s also too wordy.”

So today I tried during a writing workshop on loglines—and believe I succeeded:

With her husband deployed for the third time, a young mother must entrap the attentions of an old romantic flame in order to recapture her sense of self before it disintegrates.

The truth is, while the ChatGPT version was flashy, it wasn’t me. It contained none of the nautical themes of my story and it never clearly illustrated the protagonist’s stakes. What does she “hold dear?” Her sense of self, actually. As a Master in Creative Writing, I can quite safely say that ChatGPT failed me—and any potential agent or publisher would’ve flagged it immediately and rejected my pitch.

At the end of my work session today, I begrudgingly dragged one last essay down from my inbox for grading. I was ready to throw it all in. Give up. Hide in my bedlinens. Maybe scream. I didn’t want to do one more. One more was going to break me.

But I’m glad I did.

This final essay of the day was the most beautifully MLA-formatted essay I’ve ever seen. And it contained a semicolon use error in the first sentence. And I almost cried. The sentences that followed were so perfectly riddled with tiny imperfections that I could feel hope reinflate my chest space. This student, let’s call him Chris, was present in every single one of those sentences and errors. His voice was clear.

In my feedback to Chris, I expressed my deep gratitude that he worked so hard on the assignment. It was all I could do not to give him 100% just because it was original writing.

I’m not sure why students sign up for an online class that they don’t have time to complete.

I’m grateful for the benefits of ChatGPT and AI.

I hate the deceptive ease ChatGPT.

I am terrified of AI.

I love nothing at all more than a student’s unique writing.

I love when something new appears on the page—and I know to my marrow that it is mine.

Please, please, can we remember to relish what makes us human?

~Meg