Keep Your Eye on the Plagiarism
Citation poaching, stolen passages, borrowed work passed off as your own. There’s plenty to catch without putting somebody’s writing tools on trial.
I worked for a professor once.
Late ’70s, diving into library stacks, flipping through dozens of Reader’s Guides (books, by the way), and those infernal index cards. He’d give me a de-identified bibliography from a submitted thesis and off I went hunting for plagiarism. As part of my work, I’d pull certain graduate theses and compare the bibliographies.
Scholarly janitorial work that really helped sharpen my research skills. At this level, I wasn’t chasing the obvious paragraph thief. My job was bird-dogging citation poaching, which happens far more often than you think. Back in those days, lazy scholars looted reading lists, thereby lifting the shape of scholarship, then wore it like their own pressed shirt.
And the reason I had that job was simple. Plagiarism detection was slow, tedious, and most professors didn’t have the time to sort it all out. Some could smell plagiarism a mile away but didn’t have the hours to confirm their suspicions. So think of this as back in the day when suspicion had a budget. Or as old sailors used to say, when there’s the devil to pay and no pitch hot.
Nowadays, professors don’t need beginner researchers like me to help them pluck out the plagiarists. Really, all they need is a reliable AI tool that flags anomalies, compares lists, and catches fingerprints. All the professor has to do is what only a human can do: read the damn thing and decide if the offender really deserves to be filleted, flayed, or lightly fried with some academic probation.
And that’s a real fight. Plagiarism doesn’t just change what we can make. It changes what we can verify as truth and what trust costs when people lie for profit or sport. Or just because they’re lazy gits.
Give the Plagiarists Both Boots
Serious critics of generative AI have made good points about the weakness of AI. Model builders used material scraped without permission or compensation.
In December 2023, The New York Times filed a lawsuit against OpenAI and Microsoft with a convincing list of examples where AI-generated text matched news articles. Research presented at the Conference on Empirical Methods in Natural Language Processing showed that LLMs (large language models) can be prompted to spit out long, verbatim passages from copyrighted books used in their training data. Additional research found that AI can regurgitate niche written material, such as sections of doctoral theses or specific academic arguments. And how about that time when a New York Times writer used AI to generate a book review that plagiarized parts of a review published by The Guardian.
Alas.
This kind of malarkey is a supply chain problem with an old-style long-term fix that’s as boring as a ham and cheese sandwich: detect the plagiarist, make reparations (if published), and move on.
But there’s some displacement, too. Not the melodramatic “art is dead” but the real one where companies pocket the productivity gain and downgrade humans into “optional.” Maybe it’s really an economic fight, a class fight.
Plagiarism is definitely on the trust spectrum. When synthetic output is cheap enough, then bullshit gets cheap. But let’s not lose focus. Verification against plagiarism becomes a tax we all pay from diminished trust. We solve that problem by being vigilant about plagiarists who use other people’s work. And where possible, sue the shit out of them if they try to duck the pain of discovery.
And if someone uses AI and doesn’t proofread and check their text for plagiarism—I say let ’em have it. Think Jack Kirby’s image of Captain America’s boot busting through the window. Hell… give them both boots.
At high speed.
You Want to Talk About Class Struggle?
But this is where the Luddite AI doomsayers lose their focus: the same cost reduction that enables corporate spam also hands new capabilities and capacity to the small-time individual who is just trying to reach out with an original idea.
Like the retiree who couldn’t test one draft of her novel but can now test a hundred. Or the non-native speaker who can workshop their voice with reliable corrections. How about the novice writer with a fabulous idea but not the polish? You would deny those people the tool that equalizes their opportunity to get recognition? You want to talk about class struggle? There’s your core issue—AI as an equalizer, not just an optimizer. As I said, if you want to get real with the critique against AI, get real with your target.
My beef with “content originalists” isn’t about the concrete harms. My fight is the recognition that some of these twits think AI use ought to be a purity test for the faithful originalist. Their intention is to delegitimize whole categories of tool use.
Jack Kirby’s “Low Rent” Tropes
Save the incense. If you want to pretend that culture always ran on sealed containers of virgin ideas, go right on ahead. But you know what? Culture is recombinant. Always was and always shall be, world without end—bro.
Wordcraft cannot avoid influence. In fact, the best stories transform influence into something fresh. For example, ever notice how awesome stories subvert a familiar trope into a surprise twist? Jack Kirby pointed out that tropes are useful because they’re “low rent.” He didn’t intend this as an insult. By “low rent,” he meant that the creator could move in fast. The trope is the doorknob. You grab it to get the reader into the room, then you deliver something that isn’t just the same old furniture. That’s how creation actually works.
So what does that have to do with AI? I can ask it why my ending reads like every other revenge story. I might get a useful answer. I might get a load of crap. Maybe I’ve killed the villain when I should have let the bastard live with what he did. Now I’ve got something to work on. Asking for that critique hasn’t given me somebody else’s novel to put my name on. It’s just giving me another idea to work through.
“Tool” Equals “Guilt”?
It’s no longer what you plagiarized. No. The new equation is “tool” equals “guilt.” Just like back in Salem. Hysteria led to more than 200 people being accused of witchery. At least twenty-five innocent people died. That’s the illogic of the argument. And it’s also unethical. Tools don’t commit fraud, dammit. People do.
The moment you treat a tool as prima facie evidence of dishonesty, you aren’t defending standards. All you’ve successfully done is put on the wig and cloak of inquisition and decree stigma. Worse yet, you assume the dais of classism and wrap yourself up with the flag of ethics.
People cheat. They always cheat. They cheated when I was walking those stacks in the library. Cheaters existed before libraries existed. They’ll cheat after your “No AI” pledge fades away from memory. But you know what doesn’t fade? The damage when an accusation falsely destroys a nascent writing career. Then what can we do but “tsk” and walk on?
The Audience Doesn’t Owe You One Iota of Attention
Now we’ve opened the discussion to overlapping ethical questions about content creators who use AI. We’ll start by pointing out that all published content asks something of its audience, be it time, attention or money.
At minimum, the content creator is asking another human to surrender a portion of a finite life. But if the content creator cannot be bothered to think carefully about the work and relies on AI to slop it together, then why should anyone else spend time consuming it?
That’s where the ethics of reciprocal obligation raises its blistered hands. The creator owes the audience a reasonable degree of care. The audience doesn’t necessarily owe the creator one iota of attention. The content and the creator have to earn it.
I Sat with Jack and Talked to Him
I think human experience bestows a degree of moral authority. If I write a persuasive headline about a topic, am I obligated to admit that direct experience isn’t always required? I was in a situation where someone said I couldn’t write about Jack Kirby because I didn’t work as a comic book artist. My response: By your logic, sir, the only people qualified to write about historic events or history in general are those who had firsthand experience.
And yet, should the responsibility for content be relegated to a machine?
No…because only humans can be held responsible for what is expressed. AI might write a persuasive story about Jack Kirby’s career, but only a human can tell the story about how that career affected Jack and his family. Some topics simply need that human connection more than the polished presentation.
That’s where honesty comes in. Because only through honesty can the human acknowledge the limits of their experience and authority. I can write about Jack’s reflections on his career after he retired because I sat with him and talked to him about it. I can’t say why he went back to Marvel after he left DC. I can’t speak authoritatively about why he decided Captain America needed a round shield or why he named a monster Fin Fang Foom. But I can speculate based on what Jack said to me in person. And I can bloody well write about it, even if I use a little help from AI.
And it’s none of your goddamned business if I do or if I don’t. The ethics of care obligates us to respect the audience. The ethics of honesty obligates us to recognize what we actually know, what we have experienced, and what we are qualified to claim.
Did You Steal? Did You Lie?
The adult line is the rational standard of separating process from product. If someone plagiarizes, catch them. And oh look—with AI, you can catch them faster than ever. If someone lies about their sources, nail them.
If someone fails to post the proper claims and disclosures—be it academic, journalistic, or commercial—hold them firmly over the fire. But if someone uses AI the way competent people use tools for drafting, exploring, revising, and pressure-testing, the work is theirs—lock, stock, and barrel. Celebrate what AI gives us.
Celebrate the fact that more people (even ESL folks) can afford iteration and variation. Rejoice that the same tech that makes cheating easier can also make it easier to catch.
P.S. Disclosure for the purity police: I used AI to proofread this bit. Most modern proofreading tools are AI under the hood. Sorry if that little revelation collapses your moral system.
About: Ray Wyman, Jr is a content creator, communications professional, and author with more than 30 years of experience. Visit LinkedIN or Raywyman.com for more information.


