Regulile UE obligă Anthropic să marcheze conținutul generat de Claude. Măsura, aplicată global

Anthropic plans to implement invisible watermarks in texts generated by its AI model Claude to comply with EU transparency regulations. This marking will apply to all Claude services and will include both imperceptible patterns in the text and metadata for files. The initiative aims to enhance content traceability and prevent unauthorized modifications. The watermarks will be globally applied, not just in the EU, reflecting the influence of EU standards on global AI practices. However, the presence of a watermark only indicates potential use of Claude, not the complete history of the content.