Anthropic explains how Claude’s invisible text watermarks will work

The Verge • 8/17/2026

Summary

Anthropic has announced plans to implement invisible watermarks in text generated by Claude to comply with Europe's AI transparency rules. The text marking system is based on a version of the SynthID-Text approach, an open-source watermarking technology developed by Google DeepMind that utilizes detectable patterns from wording probabilities. This watermarking feature, along with C2PA support for images processed by Claude, is being introduced to fulfill obligations under the European Union's AI Act, which mandates that synthetic audio, image, video, and text include machine-readable marks.

Share:XRedditLinkedIn

Advertisement

Lindy Score Breakdown (V4.2)

52d
Age
1
Sources
from cluster
1250
Hours Since Seen
Final Score0/100
CategoryAntiLindy
StatusArchived
Recency Multiplier0% (0.5^1250/48)
Hero EligibleNo
Score is 0 because recency decay (0.5^1250/48 = 0.000000) reduced it below 0.5

Score BreakdownRisk 50

Source Reputation: Low-trust source (6/20 pts)
Consensus: Single source - no independent confirmation yet
Age: 52 days - proven survivor

Stories gain Lindy status through source reputation, network consensus, and time survival.

Breaking Similar stories

Anti-Lindy Similar stories