GitHub project with 11k stars strips AI watermarks from Claude, SynthID, and more
新智元 · wechat · 2026-08-17
A GitHub project called watermarks-remover has gone viral with over 11,000 stars. It removes text watermarks from Anthropic's Claude, Google's SynthID-Text, OpenAI's invisible markers, and also strips C2PA authentication data and EXIF metadata from images and PDFs. It uses a three-layer approach: cleaning invisible characters, rewriting text with a non-source model to break statistical watermarks, and stripping file metadata. It also includes pixel-level image watermark removal and MarkLLM verification. The project emerged in response to Anthropic watermarking all users to comply with EU AI Act, sparking backlash. Users report Claude refuses to remove its own watermarks, but Chinese models can do it easily.
More from Fun
- Game dev: telling Claude to "make it fast" won't get you 99th-percentile optimization — gdechichi · 2026-10-03
- Papers with 'agent' in the title get cited more: AI lit search reads titles too — rajammanabrolu · 2026-10-03
- First Model to Pass the Video Turing Test: Half of Users Thought It Was Human — Puzzleheaded-King584 · 2026-10-03
- Yann LeCun calls Anthropic CEO Dario Amodei 'deluded' and 'crazy' over cybersecurity claims — yogthos · 2026-10-03
- YouTube is filling up with AI slop that merely restates the blog posts it cites — generativist · 2026-10-03
- Every's Dev Day chat with Matthew Berman: budget gone the moment he tried Ultrafast — every · 2026-10-03