Anthropic introduces invisible watermark for AI-generated texts

Anthropic, an artificial intelligence company, has announced that it will mark texts created, edited, or corrected by its chatbot Claude with an invisible watermark. This labeling aims to ensure that the content is recognizable as AI-generated, even when copied and pasted elsewhere.

The introduction of the watermark is part of implementing new regulations from the European Union. With this move, Anthropic seeks to meet the EU's requirements that promote transparency and traceability in the use of AI texts.