Hidden AI Code Revealed in Fan-Fiction Stories
Anonymous users identified Claude-generated formatting within works hosted on Archive of Our Own.
Updated on Sept. 27, 2026 in Artificial Intelligence
Live Poll
Should authors be required to disclose if they use AI to create their stories?
On June 29, 2026, an anonymous report identified hidden AI-specific HTML code within 38 fan-fiction stories on Archive of Our Own. The findings highlighted that text copied from the Claude chatbot retains unique formatting that can be exposed by custom site skins.
Why it matters
The disclosure sparked community disputes regarding transparency, as readers maintain a strong expectation that fan-fiction is authored by people. While Archive of Our Own allows AI works, the findings underscore a growing tension between automated content generation and reader trust.
The detection method relies on a site skin that reveals hidden Claude HTML wrappers retained in rich-text editors. Researchers identified 38 works containing this code across fandoms such as Stranger Things and The Pitt.
The players
Archive of Our Own
This non-profit repository hosts fan-fiction and transformative works for an international community of writers and readers.
Organization for Transformative Works
This non-profit organization manages Archive of Our Own and advocates for the rights of fan-fiction creators.
The details
Readers identified the AI-written content by applying custom site skins that make hidden HTML tags visible. The Organization for Transformative Works, which manages the site, does not mandate disclosure for AI-generated fanworks due to concerns that detection methods are unreliable and could lead to user harassment.
Timeline
June 29, 2026: An anonymous account released a 25-page document detailing AI detection methods.
June 2026: The Organization for Transformative Works confirmed AI works remain permitted on the platform.
The Tech Race
This incident follows a pattern set by the Organization for Transformative Works policy on fanwork authorship. It marks a departure from traditional community standards by introducing automated content into a space historically defined by human-to-human exchanges.
Readers who prioritize human-authored stories may now use custom site skins to identify potential AI content, though the method is not foolproof. The lack of mandatory disclosure means users must independently verify the origin of stories if they wish to avoid AI-generated content.
The takeaway
The tension between AI efficiency and creative authenticity remains a persistent challenge for community-driven platforms. Readers are encouraged to engage with authorship transparency policies to ensure they align with their preferences for human-centered content.
Further reading
For broader context on the evolution of machine-generated content, explore our Artificial Intelligence section.
Source note: This article includes information reported by International Business Times UK.
Live Poll
Should authors be required to disclose if they use AI to create their stories?
