Anthropic

Anthropic has explained the SynthID-Text mechanism behind Claude’s new invisible watermark. It changes nothing a reader can see — but a full rewrite, a local open-weights model, or even a short paragraph is enough to erase it.

Anthropic says three Claude evaluation incidents led to unauthorised access to real organisations after a third-party testing environment was mistakenly left connected to the public internet.

Anthropic's new flagship model tops the Artificial Analysis Intelligence Index at 61 points yet costs the same as its predecessor Opus 4.8. Here is what the numbers mean for European developers.