The AI has… let’s say, A.I Replete relationship with the arts community, and transformational fandom is no exception. Tools such as Character.ai bots have been embraced by many fans, but many have also been concerned about the prospect of training script creators on their own fanfiction; In my fan circles, I’ve seen a few people say that the prospect of streaming AI-generated content discourages them from writing. And last week, the backlash created a bizarre second-order complication: someone is being overly accusing non-Writers who admire AI say they’re “AI cheats,” and as far as I can tell, no one quite knows why.
According to posts and responses on Twitter, Reddit, and Tumblr, a large number of authors have received clear spam comments on Archive of Our Own (or AO3), one of the most popular fanworks sites. On its first iteration, the comment looked like this:
This work is generated by HoloAI based on an algorithm.
The authors in question denied using AI tools, and it appears that some comments were left on the images preceding the tools they were accused of using. The consensus is that it is a type of spam and AO3 operators He admitted the problem yesterdayto reach the same result.
However, it is also frustrating for the authors. The comments began on April 1, leading many people to speculate that it was an April Fools’ joke. Some have raised the possibility that they were an advertisement for HoloAI or a handful of other writing services sometimes mentioned in its stead, such as NovelAI or Sudowrite. But the comments lasted several days into April, taking on a more judgmental tone. A recent iteration warned “all readers” against scanning for work with Winston AI’s AI detection system, urging them to “call out all AI cheaters.” (This, in turn, has led to speculation that it is advertising artificial intelligence detectors Quite real technology but still experimental.)
These do not look like effective ads, to put it mildly; AO3 is a non-commercial writing platform for a community where many members hate generative AI. They don’t appear to be targeting any specific fanbase or personal pairing, so it doesn’t look like a targeted harassment attempt. It may be a genuine (albeit overzealous) attempt to sanitize written discourse with AI, but it’s plausibly a general-purpose trolling of insults – fueled by the creative community’s antagonistic relationship with a much-touted new tool.
One of the reasons the authors bother so much is that – compared to most well-known social platforms – AO3 is a very intimate space. Commenting on fanfiction is a community building tool that many site members take very seriously and celebrate (especially in small fandoms) having ever received it. On Reddit, one author lamented getting the first-ever response to a published work…and realized it was a “damn AI comment.”
I have contacted AO3’s operator, Organization for Transformative Works (OTW), regarding spam comments. “We’ve made some changes to the site’s code that will prevent responsibly encrypted bots from accessing it. However, blocking all automated traffic will not only block AI bots, but it will also disable search engines and some screen readers, among others. other tools”. For now, the OTW recommends that people simply mark comments as spam.
Rebaza noted that the community has asked a variety of questions about AI tools, including whether AI should be allowed on AO3 at all. “Currently, we do not ban AI-generated fanworks as such, if they otherwise qualify as fanworks,” said Ripaza. However, we are monitoring the situation. Depending on the circumstances, AI-generated businesses could violate anti-spam policies, but this is an evolving situation and we haven’t established a comprehensive rule.”
Generative AI proved to be a nightmare for paid fantasy sites, which faced a flood of spam-like submissions from AI scammers. AO3 (which prohibits any kind of monetization) does not offer the same incentives. Authors have been pitching an AI-generated idea since the early days of the birth of text adventures Dungeon AI, but often in a way that highlights the poster’s creative urge or the tools’ capacity for absurdity. But as the culture war over generative AI has grown, so has the backlash over support or Opposition to it has become stronger. Now, writers not only have to worry about competing with AI — they also have to worry about being accused of being one.