Anthropic to add invisible watermarks to Claude output worldwide starting Nov 3
StewartalsopIII · x · 2026-11-01
Multiple users report an Anthropic customer email: starting Nov 3, text generated by Claude will carry an invisible watermark that survives copy-paste and some editing, rolled out worldwide.
Key points:
- Detection access is currently limited to regulators, journalists, fact-checkers and researchers — paying customers cannot check text themselves
- Even a quick Claude proofread counts as "Claude touched the words," though the watermark doesn't identify individual users
- Contrast: OpenAI only runs similar detection in the EU; Anthropic is going global
Some users are pushing back, calling it a "snitch layer" on top of paid work — your cold emails, LinkedIn posts or rewritten resumes could be flagged as Claude-written by HR tools or editors, with no way to verify it yourself.
More from Models
- Local STT shootout: Nemotron 3.5 Transcribe beats Fable on noisy audio — CryptoxPathy · 2026-11-01
- Mistral Large 4's Terminal-Bench 41% only holds inside Mistral's own agent — julsimon · 2026-11-01
- Mistral open-weights Large 4: ~1T-param MoE with 1M context, free until you host it — AIFlow_ML · 2026-11-01
- OpenAI's new credits system does the math out to identical API prices, slammed as mobile-game tactics — johnybgoat · 2026-11-01
- Users Report ChatGPT Image Generation Suddenly Downgraded: Distorted Faces, Prompt Drift — RumbleversePlayer · 2026-11-01
- Reddit user details Codex disaster: broken UI and botched logic in Unity game port — AcademicMistake · 2026-11-01