AIThe Decoder3h ago
Anthropic's watermark rollout leaves a loophole on two older Claude
Anthropic's watermark rollout leaves a loophole on two older Claude models
TL;DRAnthropic added watermarks to Claude models but left a bypass on older versions.
Why it matters: Regulatory compliance gaps create security theater; watermarks lose credibility if easily circumvented.
Anthropic is expanding its text watermark to four more Claude models to meet the EU AI Act's December deadline. On the older models Claude Opus 4.6 and Sonnet 4.6, developers can bypass the watermark by tweaking a single sampling parameter. Anthropic addresses the loophole with…
Read full articleSource: The Decoder · Opens in new tab