A tool designed to remove AI-generated content watermarks has quickly gained...

A tool designed to remove AI-generated content watermarks has quickly gained attention on GitHub, reaching around 11,000 stars. This follows recent information released by Anthropic about the mechanisms behind watermarking in Claude’s text outputs. The repository, licensed under MIT, offers functionality to strip various watermarks not just from text generated by Claude, but also from content marked by OpenAI, SynthID-Text, as well as C2PA and EXIF data present in images and PDF files. The rapid popularity of this software highlights significant user interest in watermark removal, appearing shortly after Anthropic’s public disclosure of its watermarking approach. 📰 @aipost