Watching writers get accused of using AI on the word of a detector makes me angry. These tools guess, and people are losing contracts and reputations over the guess. An author can now lose a publishing contract because a stranger uploaded their manuscript to a website and got a number back. Nobody proved a machine wrote the book. A percentage appeared on a screen, and it was high enough that the people holding the contract decided the risk was theirs to avoid.
Whatever you think about AI in publishing, that’s not how evidence works. I don’t much care what tools touched a book if the reader is engaged, and I care a great deal when somebody’s career ends on a guess. These detectors can’t do what they’re being asked to do, and you can see why from how they function, whatever anyone believes about machines and art.
The number on the box is not the number that matters
Detection companies advertise accuracy in the high nineties. Some claim better than 99 percent. Those figures are usually honest as far as they go, and they’re close to meaningless in practice.
An accuracy figure comes from running the tool against a test set built for the purpose, half human writing and half machine writing, balanced on purpose so the score means something.
Nobody uses it that way. A reader doesn’t feed a detector a random sample of ten thousand books. A reader feeds it the one book that felt wrong. Every text that reaches these tools has already passed through a human filter that selected it for being suspicious.
Watching writers get accused of using AI on the word of a detector makes me angry.
That changes the question completely. The advertised figure answers “how often is this tool right on a balanced sample.” The question that matters is “given that a suspicious-seeming manuscript was flagged, what are the odds it was machine written.” Those are different numbers, and the second one depends on how common machine-written manuscripts are among the ones people find suspicious. Nobody knows that figure, and the detection companies don’t publish it, because it’s not a property of their tool.
Nobody knows that figure, and the detection companies don’t publish it, because it isn’t a property of their tool.Share on X
Anyone who has worked with medical screening knows this problem. A test with excellent accuracy, applied to a population where the condition is rare, produces mostly false positives. The test is being asked a question it wasn’t built to answer.
That base-rate problem is first-year statistics, and the companies selling these tools know it. Advertising a balanced-sample accuracy figure to editors who’ll only ever run the tool on a manuscript they already suspect is marketing that gets writers hurt.
Why do AI detectors give different results for the same text?
Because most of them are producing an estimate, and estimates move. Multiple reported cases involve the same document, unchanged, run twice, returning materially different scores, and that alone should end the conversation. In one documented instance a set of published opinion pieces was flagged, then re-run, and produced results that didn’t match the first pass.
A measurement that doesn’t repeat is a reading. If a scale gave you a different weight each time you stood on it, you wouldn’t use it to make decisions, and you’d certainly not use it to make decisions about somebody else’s career. Formatting alone can move a score. Strip the front matter, change the line breaks, remove the chapter headings, and the number shifts. Which means the result depends partly on what was done to the file before it was uploaded, and by whom, and with what intent.
Nobody can tell you what the tool is looking at
Ask a detection company what specifically triggered a flag and you won’t get a straight answer. In at least one public exchange, the head of a detection company was asked directly and could only speculate, offering correct grammar and coherent argument as possibilities. There’s no rebuttal available. You can’t address the evidence because nobody will say what the evidence is. You’re left proving a negative to people who have already made a decision, using a number nobody can explain, produced by a process nobody will describe.
In any other context we’d call this an accusation without particulars. Courts throw those out. Publishing appears to accept them.
Half the time the people making the accusation in writing groups aren’t even using software. They’re reading for patterns. A list of three, an idea that comes in a pair, a tidy sentence, and the verdict is in.
A seventy-year-old fiction writer who taught technical subjects in the nineties described it perfectly in an essay this fall. She’s made her own lists her whole life, and now she’s a human trying to get the world to admit she isn’t a robot. She’d like a ruling by Thursday. Another writer said her journal articles from the mid-1990s get flagged as heavily AI generated, years before anyone could have used a chatbot to write them. I hate that. When a method convicts work that predates the machine, the method is broken.
The finding that should have stopped all of this
In 2023, researchers at Stanford led by Weixin Liang published a study in the journal Patterns with a title that leaves little room for interpretation: GPT detectors are biased against non-native English writers. They ran essays written by non-native English speakers through seven widely used detectors. The detectors misclassified more than 60 percent as AI generated. One detector flagged 98 percent of them. The same tools classified essays by native speakers correctly at high rates.
The mechanism isn’t mysterious once you see it. Detectors lean heavily on a property called perplexity. That’s roughly a measure of how surprising the next word is. Machine writing tends toward the predictable, because predicting the likely next word is what the machine does. Second-language writing also tends toward the predictable, because a writer working outside their first language reaches for the safe construction, the common word, the sentence pattern they’re confident in.
The detector can’t tell those apart. It was never able to. It sees low lexical variety and flags it.
This finding makes me furious. The writers most likely to get flagged are the ones who worked hardest to write well in a language they didn’t grow up speaking, and the tool punishes them for choosing the safe, correct sentence.
The same logic extends past second-language writers. Anyone writing in plain, unornamented prose is at elevated risk. So is anyone writing deep inside a genre with strong conventions. That brings us to the next problem.
What are AI detectors really detecting?
When flagged passages get published, they’re worth reading carefully, because they’re rarely what you’d expect. Phrases cited as evidence of machine authorship have included lines about a pulse thundering in someone’s ears and drowning out the crowd. Romance writers have leaned on that cliché for forty years. Anyone who calls it a machine fingerprint hasn’t read much romance, and I wouldn’t let that person judge a manuscript.
The reasoning behind flagging such a phrase is that it appears often in suspected machine writing and rarely in confirmed human writing. But consider what that comparison captures.
The models learned from mountains of genre fiction, so of course they sound like genre fiction. So does every writer who read the same books. A human writer working in that genre reproduces the same conventions, for the same reason: it’s what the form sounds like.
If a detector keys on conventional phrasing, then the more thoroughly a writer has absorbed the conventions of their genre, the more suspicious they appear. That’s an inversion of what anyone wants. It penalizes fluency in a form.
There’s a version of this critique that cuts the other way and deserves stating. Some of the books at the center of these controversies were, by wide agreement, not very good. Flat prose, tired images, dialogue that doesn’t land.
That’s a real editorial judgment and it’s worth making. It’s also a completely different claim from “a machine wrote this,” and collapsing the two is how careers get destroyed over a taste dispute.
The chain of custody nobody is asking about
Set the technology aside and look at what has to happen for one of these accusations to reach a publisher. Someone obtains a copy of the manuscript. Sometimes that copy is pirated. They may modify it, stripping the parts that would skew a result. They upload another person’s copyrighted work to a third-party service, without the author’s knowledge or consent, and that service processes it on servers the author has never heard of. Then a screenshot of the resulting number travels further and faster than any correction ever will.
Every step of that would be inadmissible anywhere that takes evidence seriously. Unknown provenance, unverified handling, an instrument that doesn’t reproduce, and an interested party operating it. In publishing it’s a Tuesday.
And there’s a commercial dimension worth naming plainly.
Every accusation is free advertising for the detection company whose logo appears on the screenshot. Search interest in these tools spikes in step with each controversy. That doesn’t mean anyone is acting in bad faith. It does mean the incentives all point one direction, and nobody involved is neutral.
What should an author do about AI detection accusations?
The uncomfortable answer is that you can’t win the argument after it starts, so the work is in what you keep beforehand.
Keep your drafts, dated. Not the final manuscript. The bad version from eighteen months ago, the one with the abandoned subplot. Cloud documents with version history are ideal because the timestamps aren’t yours to fake. A file that shows a chapter growing over six weeks is worth more than any denial.
Keep the debris. Notes, outlines, the research you did, the emails where you argued with your editor about chapter nine. The mess of a real process is difficult to manufacture after the fact and instantly recognizable to anyone who has made something.
Write down what you use and when. If you use a machine for spelling, grammar, research or brainstorming, put it in writing before anyone asks. A disclosed process is a position. An undisclosed one becomes an admission the moment somebody goes looking, whatever it was.
Get it into the contract. If you work with a publisher, a collaborator or an editor, agree in advance what tools are acceptable and who’s responsible for what. Two of the recent cases involved manuscripts that passed through another person’s hands, and the resulting blame moved in circles.
Don’t run your own work through a detector to reassure yourself. You’ll get a number, the number will mean nothing, and if it’s high you’ll have talked yourself into a panic over a reading that doesn’t reproduce.
If a publisher or editor raises it, ask in writing what evidence they have beyond a score, and keep every reply. Make them state the particulars. A vague suspicion is easy to voice on a phone call and a lot harder to defend on paper, and a writer is entitled to see what he’s accused of.
The part that should worry publishers more than authors
Look at the sequence in these cases and something odd emerges. Agents read the manuscript and championed it. Editors read it and bid on it. In one case fourteen publishers competed and the price climbed past two million dollars. Every professional in the chain read the words and wanted them. Then a number appeared, and the same professionals reversed. The manuscript hadn’t changed. Nobody learned anything new about how it was written. A screenshot arrived and they set aside the judgment of every experienced reader who had touched the book in favor of it.
That should trouble anyone in publishing far more than the possibility that a machine wrote a novel.
If professional editorial judgment collapses the moment a percentage appears, the judgment wasn’t doing much work in the first place. And the question worth asking is what editors were doing with these manuscripts, because flat prose and tired metaphors are ordinary editorial problems with ordinary editorial fixes. Marking them as suspicious instead of fixing them is a strange way to run a publishing house.
Where this leaves us
None of this is an argument that machine-written books are fine, or that nobody is passing off generated text as their own. People are, the volume is rising, and it’s a real problem for readers and writers both.
It’s an argument that the tool being used to police it doesn’t work well enough to police anything. A detector that can’t reproduce its own results, can’t explain what it measures, and misclassifies the majority of second-language writers belongs in a publishing decision about as much as a coin with a logo on it. The cost of using it anyway lands on specific people. Careers that took a decade end in an afternoon, and the correction, if it ever comes, reaches a fraction of the audience that saw the accusation.
If you’re writing a book and this frightens you, the real reassurance is small but real: keep your drafts, write down your process, and make something only you could have made. I’ve written elsewhere about whether AI can write your book and about the phrases that make writing sound machine-made, both of which are more useful than any detector. And if you want somebody in your corner who documents the whole process as a matter of course, that is what I do.
My take
No publisher should cancel a book over a detector score. The tools can’t say what they key on, the same text can score differently twice, and they flag non-native speakers at enormous rates. Keep every draft. If you’re accused, ask in writing for the evidence beyond the score. Publishers should worry more about their own judgment, since a percentage overruled every editor who’d read the book.
Source
The cases behind this piece were reported in a video essay that traces the sequence in each one and links its own sources in the description. I’ve left the authors unnamed here on purpose, since repeating an accusation with a name attached is the exact harm this article is about, but the reporting is worth watching:
Any publisher or editor who treats a detector score as proof is gambling with somebody’s livelihood on a machine that can’t tell a careful second-language writer from a chatbot, and I think that’s shameful. Ask for the evidence a human would accept. If there isn’t any, drop the accusation and retract it as publicly as it was made.
