How AI Watermarking Actually Works at the Token Level — And Why Developers Should Pay Attention
Anthropic is watermarking Claude's output at the token level. Here's how the green/red list mechanism works and what it means for developers building on LLM API
On August 11, 2026, Anthropic announced that Claude would begin embedding watermarks into the text it generates. This change came nine days after the EU AI Act, Article 50, took effect on August 2. Unlike traditional watermarks, the AI watermark operates at the token level, meaning it's embedded within the model during word selection, making it invisible to readers and undetectable without a specific cryptographic key.
Written by urgent.news from HackerNoon's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
