Claude's Invisible Watermark Cracked Within Hours, Sparking Debate on AI Watering Regulation
Anthropic added invisible machine-readable watermarks to Claude's outputs to comply with the EU AI Act, but a developer bypassed it within four hours. A former DeepMind researcher criticized regulators and media for misusing SynthID despite its well-known limitations.
2026-08-24 ~ 2026-08-24 · 2 related posts
- Claude's invisible watermarks cracked within hours; override code gets 20k bookmarks — deliprao · 2026-08-24
- Deepmind Researcher: SynthID Watermarking Limitations Known, Misused by Regulators — deliprao · 2026-08-24