Anthropic watermarks Claude outputs to satisfy EU AI Act

Anthropic now watermarks Claude's outputs with invisible code marking text as AI-generated, a policy rolled out to satisfy the EU AI Act's Transparency Code. Reddit users are split, with some calling the watermarks unethical and others supporting them as a sensible tracking method.