Watermarks-remover tool strips AI provenance from content
TL;DR. A new open-source tool, Watermarks-remover, helps users strip AI provenance and watermarks from text and files like images or PDFs. - The tool removes invisible Unicode marks, statistical text watermarks from LLMs like Claude or Gemini, and C2PA metadata. - It supports various file types including PNG, JPEG, SVG, PDF, DOCX, HTML, and Markdown. - The project aims to provide privacy and hygiene for content users own, addressing issues of AI output attribution.
- Watermarks-remover is an open-source tool to remove AI provenance marks.
- It targets multiple layers: invisible Unicode, statistical text watermarks, and C2PA/EXIF/XMP metadata.
- The tool supports various file formats and works with outputs from models like Claude, Gemini, and OpenAI.
- The project seeks to enhance privacy and hygiene for users' content by stripping AI attribution data.