Always afraid of a lot of things

I must find...the HOES.

"Chessmen XIX", by Erwin Olaf

come a little bit closer and listen to me for just a moment:

This is a bot comment. It's SPAM.

What is more, it's spam that's in favor of genAI slop. Give me a few minutes of your time and I'll explain how, since I spent a couple of hours digging into it.

Comments like this have been spotted and quickly identified as malicious spam on AO3 as early as 2023. If you have spent any time in AO3 writing spaces, then you should have run into discussion and warnings about them, because they were everywhere for a while. They got a little sneakier about it (to a given degree of sneaky, because the comment itself is still likely AI-generated as well) but the scam goes something like this:

Somebody accuses your lovingly handcrafted fic of being genAI. They used to straight up name a generator, saying something like "your fic sounds exactly like something OrphanCrushingSlopMachine-9000 would generate". The point of that probably was to use your fic as indirect advertisement for "OrphanCrushingSlopMachine-9000". They counted on somebody gullible reading the comment and deciding to check the specified AI website/tool and hopefully trying it out, because if it really could generate text like your fic then they could generate their own fics and wouldn't that be great?

They don't do this as much anymore, because it gave away the game too quickly and made the spam easier to identify and report. Now the new script starts by making some vague claims about your fic being genAI, usually with a degree of emotional manipulation tacked on (see how the comment claims that the fic was recced and it suggests that the reader is disappointed because the plot seems interesting? They skim superficial information about the fic and the chatbot adds it to the comment to make it seem more legitimate) without naming any specific generators or sites. When you take the bait and try to deny the claim or even try to give proof, that's when they will show their hand.

They will either start insisting that no, your fic really sounds EXACTLY like something put out by OrphanCrushing-SlopMachine-9000 (thus circling back to the same indirect advertisement tactic from before)

Or they will say that no, sorry, they can't take your word on it and your proof isn't good enough, but there is this one specific AI checker they trust above anything else, you should plug your fic into that and it will prove without question that your fic isn't AI.

Do you see the trap there? The "AI checker" in question is either a paid subscription (turning this into a straightforward scam) or something that runs off the same model as one of the AI generators they are supposed to be checking for. The whole point is to try and scare you into putting your authentic text into an shady AI-powered site, thus providing the LLM with more training data.

Most of the time the comments come from guest accounts, but sometimes they are from newly registered or even hacked accounts, so if you see something like this, even coming from an account that has bookmarks or even posted fics of its own, it's very likely a bot. Don't engage, just report and delete.

Fanlore has a page keeping track of all known, wide-spread AO3 comment bots. If you see a suspicious comment and you aren't sure if it's legit or not, maybe check there first before taking it at face value.

I personally haven't seen comments like these in some time (I got about half a dozen or so last year) but that doesn't mean they aren't around anymore. There seems to be a newfound push to present people opposing genAI in fanfic spaces as unstable lunatics who are looking for any excuse to accuse upstanding innocent writers of using genAI so they can freely bully them and I've certainly seen a lot of references to comments like these to prop that claim up, so I might just have been lucky to avoid them.

All the same, these comments are NOT coming from people opposing genAI. They are not the sign of some nebulous witch hunt on AO3 authors. If you've been duped by this kind of comment before: I'm sorry, I hope this post will help you identify them in the future. If you have bought into the narrative being spun about the "anti-genAI witch hunt destroying and bullying innocent writers", please consider for just a moment:

Who benefits from it if you turn your eyes away and hold your tongue even at the most blatant examples of genAI slop for fear of being branded a bully? Who has a vested interest in destroying or discrediting anyone and anything that can prove that something was made, at least partially, with genAI?