On June 29th, an anonymous account named @heatedrivalryai launched a tool designed to identify fanfiction written by artificial intelligence rather than humans. This initiative targets the Archive of Our Own, a major repository for fan-created stories, by scanning text for patterns that supposedly indicate machine generation. The creator claims the system can reliably distinguish between human and AI output, yet the detection methods remain unverified and potentially flawed. Any writer submitting work to the site risks having their story flagged or removed based on these questionable metrics.
This conflict highlights a growing tension within creative communities regarding authenticity and the role of technology in storytelling. Readers and writers have long shared informal tips to spot AI writing, but a formalised detection tool raises concerns about false positives and the potential for abuse. The situation underscores the difficulty in defining originality when generative models produce text that mimics human style.
- The tool scans for specific linguistic markers associated with large language models
- False positives could lead to the removal of legitimate human-written works
- No independent review board currently oversees the detection process
