Anthropic adds invisible watermarks to new Claude models' text

AI/Anthropic/
Anthropic adds invisible watermarks to new Claude models' text
Illustration

Anthropic says new Claude models embed an imperceptible, machine-readable watermark in generated text that survives copy-paste, while supported files carry signed C2PA provenance metadata. Marking spans the API, Claude, Claude Code and Claude Tag worldwide, driven by the EU AI Act's transparency code; older models will follow. Anthropic notes a detected mark shows Claude processed the text, not that it wrote it.

3 sources

  1. 01TechCrunch — Anthropic says it will watermark text generated by its AI modelstechcrunch.com
  2. 02The Decoder — Anthropic watermarks all Claude outputs globally with marks that may persist through some editingthe-decoder.com
  3. 03Fortune — Anthropic plans to add an invisible mark to AI text as the industry scrambles to police AI slopfortune.com