Claude will apply invisible watermarks to AI text and images
Anthropic commits to invisible watermarks on all Claude output — a first-mover compliance play that reshapes how AI-generated content gets tracked.
Why it matters
Anthropic is embedding machine-readable provenance data into Claude text and images to meet EU transparency rules. This is the first major lab to commit to invisible watermarking at scale, setting a precedent for detectability and compliance that will ripple through regulatory and industry conversations.
The key facts
12 to knowAnthropic pledging watermarks on all Claude-generated text and images
Watermarks are machine-readable but invisible to human eyes
Includes digitally signed provenance metadata where supported
Driven by EU AI transparency regulations
Future commitment, not immediate rollout
Will enable detection of Claude-generated content by platforms and individuals
Anthropic pledging machine-readable watermarks on Claude-generated text and images
Watermarks invisible to human eyes but detectable by machines
Changes driven by European AI Act transparency obligations
Digitally signed provenance metadata where supported
Future commitment, not immediate implementation
Intended to help people and platforms detect Claude-generated content
Go to the source
The Verge AItheverge.com
Publisher excerpt: Anthropic has pledged to start marking Claude-generated text and images with machine-readable data, in an effort to comply with European rules for AI transparency. "Generated text will carry embedded watermarks, and generated files will include digitally signed provenance metadata where supported,"…