AI path · course 115 of 258

The Watermark You Can't See

Beginner · 3 lessons · 0 complete

Anthropic began watermarking Claude's text output and attaching C2PA Content Credentials to supported files in August 2026, with Fable 5.1 and Mythos 5.1 confirmed and older models being added. This course covers how text watermarking differs from metadata, what C2PA actually proves, and why every service claiming to detect AI text is making an educated guess.