18 August 2026

Anthropic adds invisible watermarks to Claude text to meet EU rules

Anthropic, the company behind the Claude chatbot, is embedding invisible patterns into text Claude generates so regulators can verify it came from AI, required by the European Union's AI Act.
The watermark works by having Claude make arbitrary choices between similar words (like 'overcast' versus 'grey') guided by a hidden key, creating a detectable pattern that readers cannot see.
Anthropic says watermarking does not change output quality, cost, or function, and light editing will not remove it, though a complete rewrite of the text will.
Generated code will carry weaker watermarks because the model has fewer word choices when writing functional code, though comments within code can still be watermarked.
Google's Gemini chatbot has already used the same SynthID-Text watermarking approach since 2024, and other major AI developers must also implement watermarks under EU law.