Claude now watermarks text and adds C2PA metadata to media, complying with the EU AI Act's Article 50 transparency rules.

Invisible watermarks will be woven directly into text generated by Claude models. It will not alter the meaning, quality, or readability of a response and can travel with the text when copied ...

Anthropic will embed hidden marks in Claude-generated text and attach digital origin records to some files, though it ...

How will Anthropic mark Claude-generated content? What does the EU’s AI Act have to do with it? And to what extent will these measures help distinguish human writing from synthetic text?

Anthropic is adding invisible, machine-readable watermarks to Claude-generated text and provenance metadata to supported ...