Someone keeps accusing fanfiction authors of writing their fic with AI, and nobody knows why

Artificial intelligence has… let’s say, a fraught relationship with the arts community, and transformative fandom is no exception. Many fans have embraced some tools like Character.ai bots, but plenty have also bristled at the prospect of text generators getting trained on their fanfiction; in my fandom circles, I’ve seen a few people say the prospect of an AI-created content flood discourages them from writing. And in the past week, the backlash has created an odd second-order complication: someone is rampantly accusing non-AI fan writers of being “AI-using cheaters,” and as far as I can tell, nobody quite knows why.

According to posts and responses on Twitter, Reddit, and Tumblr, a slew of authors have been receiving apparent spam comments on Archive of Our Own (or AO3), one of the most popular fanworks websites. In its earliest iteration, the comment looked like this:

This work has been generated by HoloAI based on an algorithm.

The authors in question have denied using AI tools, and some comments were seemingly left on fics that predate the tools they’re accused of using. The consensus is that it’s a type of spam, and AO3’s operators acknowledged the problem yesterday, coming to the same conclusion.

It’s also, however, frustrating to authors. The comments started around April 1st, leading many people to speculate they were an April Fools’ joke. Some raised the possibility they were an ad for HoloAI or a handful of other writing services that were sometimes mentioned in lieu of it, like NovelAI or Sudowrite. But the comments have continued several days into April, and they’ve acquired a more judgmental tone. One recent iteration warned “all readers” to scan the work with the Winston AI detection system, exhorting them to “call out all AI-using cheaters.” (This has, in turn, led to speculation that it’s advertising AI detectors, a technology that’s quite real but still experimental.)

These don’t seem like effective ads, to put it mildly; AO3 is a noncommercial writing platform for a community where many members loathe generative AI. They don’t seem aimed at any particular fandom or character pairing, so they don’t look like a targeted harassment attempt. It might be a genuine (if overzealous) attempt to purge AI-written fic, but it’s most plausibly a kind of general-purpose insult trolling — fueled by the creative community’s antagonistic relationship with a much-hyped new tool.

One reason this bothers authors so much is that — compared to most well-known social platforms — AO3 is a very intimate space. Comments on fanfiction are a community-building tool that many site members take very seriously and (especially in small fandoms) celebrate receiving at all. On Reddit, one author lamented getting their first-ever response to a posted work… and realizing that it was “that fucking AI comment.”

I contacted AO3’s operator, the Organization for Transformative Works (OTW), about the spam comments. “We’ve made some changes to the site code that will prevent responsibly coded bots from accessing it. However, preventing all automated traffic would not only block AI bots, but it would break search engines and some screen readers, among other tools,” communications director Claudia Rebaza wrote in a statement from the OTW support team. Right now, OTW recommends people simply mark the comments as spam.

Rebaza noted that the community has asked a variety of questions about AI tools, including whether AI fic should be allowed on AO3 at all. “At the moment, we do not prohibit AI-generated fanworks as such, if they otherwise qualify as fanworks,” Rebaza said. “However, we are monitoring the situation. Depending on the circumstances, AI-generated works could violate our anti-spam policies, but this is a developing situation and we have not made a blanket rule.”

Generative AI has proven a nightmare for paid fiction sites, which have faced a flood of spam-like submissions from AI hustlers. AO3 (which bans any kind of monetization) doesn’t offer the same incentives. Authors have been posting AI-generated fic since the early days of text adventure generator AI Dungeon, but often in a way that highlights the poster’s creative prompting or the tools’ capacity for absurdity. But as a culture war has grown around generative AI, the backlash for supporting or opposing it has gotten stronger. Now, writers don’t just have to worry about competing with an AI — they have to worry about getting accused of being one, too.