Anthropic explains how Claude's text watermarking works
surprisetalk · hn · 2026-08-15
Anthropic published a blog post detailing the mechanism behind Claude's text watermarking. The technique embeds a verifiable signal by statistically adjusting token sampling probabilities with minimal impact on text quality. The post highlights its effectiveness in identifying AI-generated content while discussing trade-offs regarding output quality and robustness challenges.
More from Models
- 'Small-Claude-shame': telling Opus 4.7 it's OK to know who you may ease it — repligate · 2026-10-16
- Bridgewater fine-tunes Qwen3-235B with Tinker, beats GPT-5.5 at 1/14th the cost — simonguozirui · 2026-10-16
- ChatGPT voice selection breaks after GPT-Live-1.1 rollout on Android — shaunralston · 2026-10-16
- Musk teases two-way Grok x Optimus integration: Grok in the bot, bot in Grok — beffjezos · 2026-10-16
- vLLM officially supports Muse Spark Open: multimodal, 1M context, agentic — vllm_project · 2026-10-16
- Liquid AI ships LFM2.5-24B-A2B: 24B MoE with 2B active params for local agents — JosephJacks_ · 2026-10-16