Anthropic extends invisible text watermark to Claude Opus and Sonnet models globally

·2026.10.31 05:48

Key point

The watermark applies to every model released since August, including Mythos 5, Opus 4.7, Opus 4.6, and Sonnet 4.6, affecting users worldwide.

Anthropic extends its invisible text watermark to Mythos 5, Opus 4.7, Opus 4.6, and Sonnet 4.6 starting November 3, applying to all models released since August.
The policy is global and mandatory for paying users, with no option to opt out.
Detection access is restricted to a private preview for regulators, media, researchers, educational organizations, and enterprises with compliance obligations, leaving general users unable to verify their own text.
The watermark relies on statistical patterns in word choice, which the article notes can be removed by rephrasing through another model, potentially allowing bad actors to bypass it while honest users remain flagged.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.