The Fanfiction Community’s Digital Witch Hunt Against AI

The tool, released by an anonymous account known as @heatedrivalryai, functions by detecting a specific tag—font-claude-response-body—that Claude injects into text when copied directly into a content management system. While verification tests confirm the code reliably flags direct pastes from the chatbot, the technical scope is narrow. Writers who draft in external processors like Google Docs or Microsoft Word effectively scrub these digital fingerprints, and the tool fails to distinguish between a fully generated story and a human author using AI for minor spell-checking or translation.

Despite these limitations, the fanfiction community has pivoted toward aggressive policing. Users are mobilizing to name and shame writers whose work triggers the red screen, operating on the belief that any AI involvement constitutes a betrayal of the collaborative, human-centric nature of fandom. This environment has already produced collateral damage; at least one author was flagged simply because a trusted editor used AI tools without their explicit knowledge.

Technological solutions for identifying synthetic text remain elusive, as major AI developers like OpenAI and Google have yet to produce reliable detection methods for copy-pasted content. For now, the community continues to rely on subjective "vibes" and stylistic critique, often misidentifying human quirks as machine-generated output. As the pressure to purge AI grows, the reliance on such blunt instruments threatens to alienate writers who prioritize the community's established tagging system over the fear of being wrongly outed by a browser script.