←── back to feed
/topics/claude-watermarking-and-text-authenticity-features

Claude watermarking and text authenticity features

2 items1 sourcesupdated 29d agotrend 0

Anthropic has added watermarking to Claude to mark AI-generated text, though a separate study found readers cannot reliably identify watermarked content. The feature aims to improve text authenticity tracking as AI-generated content becomes harder to distinguish from human writing.

  • Anthropic integrated watermarking into Claude to label AI-generated text
  • Watermarking applied to text authenticity and provenance tracking
  • Reader study showed humans cannot reliably detect watermarked AI text
  • Feature addresses growing difficulty distinguishing AI from human-written content