Models launched on or after 2 August will support machine-readable content marking.
Anthropic said its new models will begin watermarking content in the EU, as strict obligations from the bloc’s landmark legislation on AI start taking effect.
New transparency rules under the EU AI Act began taking effect earlier this month, which require certain AI systems to tell users when they are interacting with AI and how the content they are consuming is generated or altered by it.
The Claude-marker is one of more than 200 signatories of the Act’s Code of Practice on Transparency of AI-Generated Content – a voluntary tool aimed at helping businesses comply with the sweeping law. Several major AI providers, including OpenAI, France’s Mistral, Meta and Microsoft are also signatories.
Anthropic said that models launched on or after 2 August will support machine-readable content marking, meaning files generated by its AI will include digitally signed provenance metadata such as embedded watermarks on text and Coalition for Content Provenance and Authenticity (C2PA).
