Generative AI is already beginning to make its mark on schools and workplaces. However, one of the groups in the heat of the AI debate is the arts community. That is the writers, graphic designers, and illustrators who dedicate their hearts and minds to their craft.
With popularly emerging tools like ChatGPT and Bing Image Creator at everyone’s fingertips, the prospect of AI-generated content replacing human-made art and stories is a valid concern. It’s an issue that goes hand in hand with a new frustration: creators being falsely accused of using AI. Such a phenomenon has recently emerged on the popular fan fiction site, Archive of Our Own (AO3), prodding more confusion among readers and writers alike.
April Fools?
Beginning April 1st, fan fiction writers on AO3 started receiving comments stating that their works were detected as AI-created based on an algorithm. The authors then took to Reddit and Twitter to complain about these spam accusations posted on their non-AI work. At first, users speculated that the comments were some form of an April Fool’s joke. However, as the comments continued to appear well into the next week, more questions arose about the source of the spam.
One vital clue comes from a recent iteration of the accusatory comments, prompting all readers to weed out AI-using cheaters using the Winston AI detection system. This indicates that the consistent spam comments are not the work of one dedicated internet troll but perhaps a twisted marketing ploy for AI detection technologies. Thus, it appears that the blossoming world of AI continues to be a double-edged sword. The more that generative AI takes hold in society, the more we’ll be in want of detection services to help us try and determine whether a human mind is behind what we see and read.
AO3 Responds
Following the social media complaints made by authors denying the accusatory comments, AO3’s Twitter account addressed the ongoing situation, deeming them a form of spam.
Given that some of the comments appeared on fan fiction works that predate the AI technology in question, the legitimacy of the claims was quickly dashed among users. That said, the incident itself touched some vulnerable points among the fan fiction community. The advent of generative AI is disheartening to writers, whose motivation to create stories on the platform withers at the idea of having to compete with a future mass of AI-generated fan fiction.
Fan Fiction and AI
As Twitter users pointed out in the comments, AO3 still allows AI-generated content to be posted on their platform. For now, they haven’t yet established a blanket rule prohibiting AI-generated fanworks, but they are actively monitoring the situation moving forward. Discussion about how AI works could be properly tagged or flagged on the site was also introduced in the unfolding Twitter thread. However, it’s dubious that any system would be completely foolproof and hardworking creators may be snubbed through the inevitable trial and error.
On the whole, fan fiction has maintained a shaky relationship with AI technology as of late. On the one hand, AI features like Character.ai grew in popularity last year – providing a new level of parasocial engagement for fans. While a simulated conversation with a fictional character or celebrity can be a harmless exercise that helps inspire fan fiction writers, the idea of using generative AI to write entire stories is still met with general hostility. Thus, AI’s promise and potential threat continue to be at the forefront of fan fiction communities, procuring a new and tricky balance between the integrity of the writing craft and innovative avenues of technological engagement.
Click here for more about AI and the future of the audiobook industry.