Wire · regulatory
Anthropic Starts Watermarking Claude AI Text To Comply With EU AI Act Article 50(2)
◆ Sectors
◆ Geography
◆ Source
◆ Verified
Fusion42 · 23 August 2026 · Fusion42 review
Anthropic has started embedding hidden watermarks in Claude-generated texts to comply with the EU AI Act, enabling detection of AI involvement even after edits, translation, or summarisation. This watermarking approach is designed to be undetectable to readers, persist through text modifications, and comply with the EU Transparency Code requirements.
This Wire brief sits within Fusion42's coverage of AI & ML, and 13 sources have reported it between 11 Aug 2026 and 18 Aug 2026.
◆ ◆ The Wire takeaway
Watermarking AI-generated text is now mandatory in the EU, creating a compliance hurdle that founders building AI tools for European markets need to solve immediately or lose access. You must build or integrate detectable watermarking to avoid regulatory exclusion.
◆ Coverage
13 sources · first reported 11 Aug 2026 · latest 18 Aug 2026
◆ Related on Wire
◆ Topics