Anthropic has announced that it will soon add machine-readable watermarks to content processed by its AI models, not just content they generate. The move aligns with the European Union’s AI Act, which requires watermarking of AI-generated or manipulated media, with new models covered immediately and older ones granted until December 2026.
Anthropic says the watermarks will be applied globally to all new models, embedded invisibly in text and added as digital provenance metadata to other file types. The company is going beyond EU requirements by watermarking even lightly edited content, such as grammar corrections, which the EU does not consider substantial alteration.
Comments