Coders Say They Already Found Workarounds to Claude’s Invisible Watermarks

Within hours of Anthropic confirming that Claude models embed invisible, machine-readable watermarks, developer Guillaume Meyer released code that strips those watermarks by generating multiple rewrites with a non-watermarking LLM, prompting over 100 contributors and 20,000 GitHub bookmarks. The EU AI Act requires model providers to label synthetic content, but independent circumvention tools remain legal, and Meyer argues watermarks risk false positives and could degrade output quality.