Tech
Anthropic Adds Invisible Watermarks to Claude Content
By Krish
•12 hours ago
•Updated August 11, 2026
•2 min read
Size:
Key Highlights
- 1Supported Claude models will embed invisible marks in generated text and signed provenance metadata in supported files.
- 2Anthropic says the system is linked to EU AI Act transparency requirements and the marks can sometimes survive copying or editing.
- 3Detection tools are still being developed, and Anthropic says a detected mark is not definitive proof that Claude created the original content.
✓Fact-Checked by Krihaa Editorial
Share:
Comments (0)
Join the conversation
Sign in to post comments, like, and reply
📢 SPONSORED ADVERTISEMENT
India’s trusted news app ❤️
krihaa