Notes on watermarks & provenance
Short, sourced explanations — what providers actually deploy, what the standards actually prove, and where the marketing ends.
AI Watermark Checker vs AI Detector: What's the Difference?
AI detectors guess from writing style. Watermark checkers verify signals that providers deliberately embed. Confusing the two produces false accusations — here's how to tell them apart.
Does Claude Watermark Text? What Anthropic's 2026 Marking Actually Means
Yes — since August 2, 2026, new Claude models embed an imperceptible watermark in generated text and sign generated files with C2PA. Here's exactly what that does and doesn't mean.
How AI Text Watermarks Work: Statistical Marks, Hidden Characters & Provenance
AI text watermarks don't hide characters in your text — they shape which words the model picks. A plain-language explanation of statistical watermarking, and why you can't see it.
What Are C2PA Content Credentials and What Do They Actually Prove?
Content Credentials are cryptographically signed provenance records that travel with a file. What a valid signature proves, what a missing one doesn't, and how to verify one in your browser.