Everything Claude writes for you now carries a hidden mark you can't see. Anthropic started weaving an invisible signal into Claude's responses, nudging which words the model picks in a pattern only a machine can read. It stays even when you copy the text somewhere else, and generated files get a separate signed signature. They turned it on worldwide to meet the EU's new AI law. Two things nobody mentions. First, you can't check the mark yet, the detector isn't out. Second, it only shows the text was produced by Claude, not that you didn't write it, and if you rewrite it in your own words, the mark fades. I share more about AI in my community, link in bio. #ai #claude #anthropic #ainews #euaiact - @yury.ai