Anthropic starts embedding token-level watermarks in Claude's text
On August 11, 2026, Anthropic confirmed Claude will embed machine-readable watermarks in generated text. The move follows Article 50 of the EU AI Act, in force since August 2, with penalties up to €15 million or 3% of global turnover. The method is based on Google DeepMind's SynthID-Text.
- Anthropic watermarks Claude output starting August 11, 2026
- EU AI Act Article 50 mandates machine-readable AI content marking
- Penalties reach €15 million or 3% of global annual turnover
- The technique is based on Google DeepMind's SynthID-Text
Read next
AI