Are Readers Confusing Cliche Writing for AI?

A little while ago, there was a story making the rounds. This one had started on Reddit, perhaps part of the “AITA” type posts. The poster was seeking advice, specifically of the legal variety, because they had been threatened with a lawsuit.

Why was an author looking to sue a seemingly random reader? Well, it seemed this person had accused them of using AI in their writing.

So the Story Goes …

The reader was seeking advice on how they could either avoid being sued or win their case. They stated they were a “poor grad student,” who didn’t have any money to be doing things like hiring a lawyer or paying damages to an author.

This particular poster asked the community if the fact they’d pirated the book would make things better. After all, they hadn’t even paid for their copy of the book! They’d just randomly downloaded a stolen copy and read it, then decided to leave a review making accusations of AI.

Obviously, the answer there was a resounding no. The fact that you pirated the book, in fact, makes things worse. The AI accusations might be neither here nor there, but the fact you committed copyright infringement is kind of incriminating. Even if the judge doesn’t care about the fact that you made some accusations about AI use in the book, they might care that you, uh, stole the fucking book.

As an aside, I think it might be time to bring back those FBI warnings that used to show up before movies. You know, the whole “if you make copies, you’re in violation of federal law and we will find you” thing. While I don’t know that those messages ever deterred pirates, I do think they might have answered the poster’s “will the fact I pirated the book get me out of legal trouble?” question. Although, that might have meant this whole situation was less funny than it is.

Is Saying Someone Used AI Libel?

Pirating aside, the issue at hand here is likely libel. The reader left a review that said the author had used AI in their writing. They provided a few examples of what they considered “obvious” examples of AI writing in the text. (I’ll dissect that in a minute.)
The question is whether the author could win such a suit. And I do think that an author could sue someone over saying they used AI when they didn’t.

AI writing, while being championed by some, is generally not well received among either readers or other writers. While some people see uses for AI in the writing process, others eschew it altogether. Polls show that readers have poor opinions of AI-generated text in general. While some believe using AI to brainstorm is okay, most are fairly negative towards using genAI to generate the actual text itself.

So, yes, saying someone used AI to write their book when they didn’t could be considered libel. It’s likely to have a negative impact on their reputation as an author, since there’s such a large swath of people who believe using AI means you’re … not really a writer at all. If people know authors are using AI, then that could lower people’s opinions of that author’s work. It could also impact their sales and damage their reputation and their livelihood.

That is precisely the definition of libel. So, yes, accusing someone of using AI in their writing is not a claim that should be made lightly, particularly in the United States.

The Burden of Proof

In most jurisdictions, the burden of proof in libel/slander cases rests on the party claiming they were defamed. They have to prove what was said is not true. This limits the number of frivolous libel cases. In short, it keeps people from suing the shit out of each other every time something remotely negative is said about them.

The reasoning behind this is that people need to be able to be free to express opinions. So, if someone says something a little mean, that’s really okay—provided it doesn’t truly harm someone else. If you say someone is embezzling from their employer, you best believe you’re going to be hauled up in court. That kind of accusation has tangible impacts on the person you’re talking about. If you say so-and-so is kind of a bitch, then you shouldn’t be sued over it.

There are other protections around everyone suing the pants off each other for saying something negative about other people. One is that you can’t libel the dead; their reputation doesn’t matter any longer. Another is that opinion can be protected—see saying someone is kind of a bitch. Nobody can actually sue you over that, at least not successfully. It’s an opinion, which other people may or may not agree with.

This is where it is also important to establish the truth of a matter. It’s much easier to prove that someone is or isn’t a white-collar criminal engaged in embezzlement. It’s much more difficult to prove someone is or isn’t “kind of a bitch.” I mean, the fact that someone is suing you over calling them “kind of a bitch” might be a bit of a hint, but the courts don’t really work like that.

The Difference in the US

In the US, however, the burden of proving this falls on the defendant. That means the person who makes the libellous statement has to prove that what they said is true. If you can’t back your claims up, then you’d better keep your goddamn mouth shut.

This is precisely why most other jurisdictions don’t work this way. It’s easier for a plaintiff to sue for—and win—a libel/slander suit in the US than it is in, say, Canada, because you, as the defendant, have to prove what you said was true. Most people aren’t thinking about needing to prove what they say when they state something libellous. This legal framework can have the effect of chilling free speech, discouraging people from speaking out against others who are doing wrong, for fear of being sued and needing to defend their statements in court.

That’s clearly the case with our reader here. They opened their mouth and, although they provided what they felt was evidence of clear AI use in the work, they didn’t think they were going to go to court over it. They should have simply kept their thoughts to themselves.

The Evidence of AI Use Was … Cliche

They really should have thought twice about posting their review, “calling out” an author they thought was using AI, particularly because their “evidence” was scanty, at best. They called attention to phrases like “let out a breath [they] didn’t know [they] were holding” and other “hallmarks” of AI writing.

Anyone who has ever read a YA novel has encountered this phrase. Basically, all of the evidence the reader gave as being proof of the use of AI was simply cliche language.

That makes sense, to some degree. AI is trained on plenty of books, most of them pirated, some of them not, and it’s essentially a text prediction tool. It predicts the word most likely to follow. So, if we start a sentence with “She let out a breath …,” an AI tool is likely to put “she didn’t know she was holding,” because that phrase is so commonly used.

So, is using cliche language a hallmark of AI writing? Absolutely not, because those things are cliche because they’re used so much. They predate AI. Humans wrote them, and will continue to write them, and AI is simply mimicking them.

And that leads me to the next question …

Did This Actually Happen?

I’m having trouble believing a grad student who was clearly doing some pleasure reading had never encountered these phrases, that they clocked them as “clearly being AI” more than simply “the author’s writing is pretty cliche.”

Think about it. A grad student is anywhere from 22 and up. They have likely completed a whole undergrad before now. When this story was going around in 2025, genAI had been on the scene since late 2022—approximately three years, not quite. Today’s high school kids might not know the difference, but a grad student likely had some familiarity with how human beings wrote books pre-2022.

If this person is a pleasure reader, there’s a good chance they were a pleasure reader when they were in high school and undergrad—a reading career that largely predates AI. It would only be in the last few years that genAI writing would even be a concern.

We Will See Sensationalist Stories

It’s absolutely not unheard of for people to be accused of using AI—I’ve seen it happen plenty to artists, unfortunately. Usually, these attacks are led on social media—not necessarily in a review. On social media, such dogpiles can become difficult to pinpoint. As much as troll accounts definitely exist on platforms like GoodReads, it’s much easier to single out a reviewer than it is to isolate who “started” the drama on social media. This becomes even more difficult with sock puppetry, which in turn makes identifying a “ring leader” less possible. To that end, it’s surprising this person didn’t bother taking to social media, where it would be easier to hide.

While I imagine we will eventually see someone get sued over making the libellous statement that an author or artist used AI, I think we’re also going to see plenty of stories that are sold to us for sensationalism—which might be the category this one ultimately falls into.