Anthropic models will soon inject watermarks identifying AI-generated text

Anthropic announces plans to add watermarks to text generated by its models to enable identification of AI-produced content.