EU Rules Force Anthropic to Expose AI Writing Worldwide

The EU’s transparency framework is driving Anthropic to label AI-generated writing produced by Claude for users worldwide. Under the rules, the company is rolling out machine-readable, watermark-style signals embedded in text so platforms and third parties can detect or trace AI-authored content. The approach is designed to persist through copy-paste and some edits, although heavy editing or short excerpts can reduce reliability. Anthropic’s move is aimed at meeting compliance deadlines as regulators increase enforcement of generative-AI disclosure standards.