Claude text now carries a hidden mark, with older models joining on 16 November
Anthropic adds an invisible statistical mark to the words Claude writes, under its pledge to the EU code of practice.
Anthropic puts a hidden mark in the text Claude writes, so officials can tell an AI wrote it. The mark holds no information about you.
Hype check
The noise matches the real change.
Our editorial opinion, based on the sources below. How we score
Your call: hype or real?
Real = it will matter in daily use. Hype = louder than it is. One vote, and it can’t be changed.
Our view and readers’ views are opinions, not statements of fact about a product.
Why it matters
Text written or edited by Claude may be recognised later, and the mark holds no data about you.
Watch: a regulatory trend spreading across AI companies.
The details
- The mark is invisible, a pattern in word choice, and similar in idea to Google’s SynthID.
- It has run since 3 November on Claude Mythos 5, Opus 4.7 and Sonnet 4.6.
- On 16 November Opus 4.6, Opus 4.5, Sonnet 4.5 and Haiku 4.5 follow.
- Anthropic says the mark holds no information about the user, their organisation or their conversations.
- Detection is in a private preview for regulators, researchers and others, by application. A mark does not prove authorship, and short or heavily rewritten text may not carry one.
Receipts
One trusted outlet reports it. Not confirmed yet. Open the sources and check us.
Written with the help of AI tools and checked against the sources above. How we work
See a mistake? Tell us and we’ll fix it.
