Anthropic staff explains Claude's watermarking mechanism and detection status
HarperSCarroll · x · 2026-08-27
Claude's "invisible" watermark works by shifting the probability of consecutive output tokens, embedding the signal in the chosen words. Consequently, copying and pasting Claude's output retains the watermark, whereas using it for proofreading or minor edits preserves less of the mark. As of now, the technology to detect Claude's watermark has not been publicly released.
More from Safety
- ASML CFO says US Match Act restricting DUV sales to China would backfire — pstAsiatech · 2026-10-15
- Rumor: New Anthropic model imminent, apparently behind a batch of high-severity Chrome CVEs — matthew_d_green · 2026-10-15
- Fully-automated combat drones are already on the battlefield, warns Toby Walsh — TobyWalsh · 2026-10-15
- Pope's AI adviser warns Australia's copyright changes risk a new form of colonialism — TobyWalsh · 2026-10-15
- Jamie Dimon Warns of AI Agent Risks to Bank Accounts; JPMorgan Blocks Outside Agents — AccBalanced · 2026-10-15
- 'A lot of you are into alignment faking': meme nod to Anthropic's alignment faking paper — dejavucoder · 2026-10-15