Self-publishing

An AO3 Clone Sold AI Audiobooks of Scraped Fanfiction. Authors Shut It Down in a Day

A copycat of Archive of Our Own scraped fanfiction into paid AI audiobooks and closed after 2,000 takedowns in a day. What authors should learn from it.

4 November 20265 min readBy the editorial team
Pages drifting from library shelves toward a glowing server rack in a dark archive

In short

While Archive of Our Own (AO3) was offline after a security incident, a copycat site called ao3.life scraped its stories, offered them as EPUB downloads and sold AI-narrated audiobooks of them without the authors' consent. Writers answered with more than 2,000 takedown requests in a single day, and the site went dark on 2 November.

  • The clone drew 126,592 unique users in its first 24 hours
  • It was built with a swarm of AI agents running on 40 virtual machines
  • Only 11 audiobooks were sold before the 17-year-old creator pulled the plug

The fan-writing world had a strange weekend. With AO3 down, a near-identical site appeared, filled with stories its authors never agreed to share there, and started charging for machine-read versions of them. 404 Media reports that the creator, a 17-year-old developer in India who goes by Ganesh, used "a swarm of AI Agents controlling 40 virtual machines" to fill the platform. Kotaku and Gizmodo covered the backlash that followed.

The episode shows how cheap it has become to copy a whole library, wrap it in a storefront and add AI audio on top. It also shows the opposite lesson for anyone building legitimate reading or audio products: consent, rights checks and takedown handling are not legal paperwork to bolt on later. Publishers and author collectives that commission end-to-end software development for their own platforms now need those flows designed into the product from the first sprint.

For individual writers the questions are more practical. Was my work taken, and what can I do the next time a site like this appears? Here is what we know.

What happened to AO3?

Archive of Our Own, run by the non-profit Organization for Transformative Works (OTW), hosts roughly 18.5 million works, according to 404 Media. It went offline after a security incident, with downtime that could last up to four weeks. That outage created a gap: millions of readers suddenly had nowhere to go, and authors could not log in to use the archive's lock feature, which hides works from people without an account.

ao3.life moved into that gap. It copied cached versions of stories and presented them under a familiar-looking name, which was enough to pull in a six-figure audience within a day.

How did the clone work?

  1. Scraping at scale. AI agents on dozens of virtual machines collected stories, something that would have taken a team weeks only a few years ago.
  2. Repackaging. Works were offered as EPUB files for e-readers, and as AI-generated audiobooks behind a paywall.
  3. A revenue pitch. The site claimed authors would get 91 percent of audiobook income, with 9 percent kept by ao3.life, according to Kotaku. Nobody had agreed to that deal.
  4. Email requests. Kotaku reports that authors were asked to write in from their AO3-linked addresses to get paid or to have works removed, which many readers saw as a phishing and doxxing risk.
TipNever send account details or personal email to a site that hosts your work without permission. Use the platform's official takedown address or a formal DMCA notice instead.

Why did it fall so fast?

Fan writers are organised, loud and very good at spreading the word. Within a day the creator said he had "facilitated over 2,000 takedown requests." In a farewell note quoted by Gizmodo he wrote: "It was never my intention to hurt so many authors." 404 Media adds that server and bandwidth costs were heavy, and that only 11 audiobooks had been sold.

The OTW told Kotaku it was aware of the site and pursuing legal action, while noting that "the legal system is much slower than online life." In this case collective pressure worked faster than the courts. That will not always be true, and a quieter clone run by someone less willing to apologise could last much longer.

What can authors do now?

Whether you write fanfiction, serials or full novels, a few habits make the next copycat less damaging:

  • Keep your own canonical copy. A personal author website gives readers one official place to find your work and makes it easier to prove which version came first.
  • Know the takedown route. The US Copyright Office DMCA page explains how notices work; keep a template ready.
  • Watch your audio rights. AI narration is now cheap enough that anyone can turn text into an audiobook. Our note on Kindle+ and AI-narrated audiobooks covers the legitimate version of that trend.
  • Follow the policy fight. Governments are still deciding how AI training and copying should work; see our report on Australia's proposed AI opt-out.

ao3.life lasted about a day. The tools that built it are not going away, so the best defence remains the same: know where your work lives, and make it easy for readers to find the real thing.