Anthropic to watermark text from older Claude models
The move is tied to the EU AI Act, and some users argue that checks on their own writing are unfair.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
Anthropic is expanding text watermarking to older Claude versions, starting November 3, including Mythos 5, Opus 4.7, Opus 4.6 and Sonnet 4.6. The watermark shows whether a text was made with AI. The change is linked to the EU AI Act.
Why it matters
The watermarking system has proven controversial with users, who argue it is unfair if people can check whether they actually wrote their own words.
Timeline
- Nov 1, 11:59 AM ET: Anthropic expanded text watermarking to older Claude versions, to begin November 3.
- Nov 2, 10:13 AM ET: Users criticized the watermarking system, arguing it is unfair if people can check whether they wrote their own words.
- From November 3: The watermark begins applying to Mythos 5, Opus 4.7, Opus 4.6 and Sonnet 4.6.
Coverage
- Anthropic’s Claude will now leave a trail that shows whether a text was made with AI (the-independent.com)
More from Technology & AI
- China puts humanoid robots to work serving food and sorting parcels
- Newsom proposes sweeping AI safety rules for Washington
- Google releases Gemini 4 Neon AI model with million-token context
- Microsoft and Intel widen their cloud chip partnership
- Google DeepMind rolls out Genie 4 to select subscribers
- AI helped AMD shorten chip design work, OpenAI says