I created a viral AI watermark remover. I wasn't ready for all the attention.

Live updates

1. 23d ago

   Developers bypass AI text watermarks as tools go viral

   Coders have developed workarounds to remove invisible watermarks from AI-generated text, including those used by Anthropic's Claude. This follows the creation of a viral AI watermark remover that gained unexpected attention. These developments suggest a technical struggle between AI companies attempting to label machine-generated content and users seeking to erase those markers. While companies implement these systems to identify AI writing, the emergence of effective removal tools challenges the reliability of these detection methods.

   Why it matters

   AI companies use watermarking to distinguish human writing from machine output. This is intended to prevent misinformation and academic dishonesty. The ability to strip these marks undermines the primary mechanism for AI content attribution.

   Still unconfirmed

   • Coders have found workarounds to Claude's invisible watermarks.
   • A viral AI watermark remover was created by a developer who was unprepared for the resulting attention.

   What to watch next

   • Evidence of widespread use of watermark removal tools in academic or professional settings.
   • Updates from Anthropic regarding the efficacy of Claude's invisible watermarks.
   Sources used for this update (5)

   1. 404 Media — Anthropic’s Text Watermarking Proves AI Companies Do Not Care at All About Writing
   2. Business Insider — I created a viral AI watermark remover. I wasn't ready for all the attention.
   3. WIRED — Coders Say They Already Found Workarounds to Claude’s Invisible Watermarks
   4. The New Yorker — The Mark of the Machine
   5. CNN — What does it mean to put a ‘watermark’ on AI text?
   confidence 70%