Anthropic explains how Claude’s invisible text watermarks will work

Anthropic has announced that it will use a version of Google DeepMind's open-source SynthID-Text system to apply invisible watermarks to text generated by Claude. This initiative is designed to comply with the European Union's AI Act, which mandates machine-readable markings for synthetic content. The watermarking technique subtly influences low-stakes word choices based on a secure key, leaving a detectable pattern without affecting the quality, content, or cost of Claude's outputs. Other major AI developers, such as Google with Gemini, are also implementing similar transparency measures.

(Source：The Verge)