Living definition
What is Claude Text Watermarking?
Privacy-Enhancing Technologies & Content Provenance Security & Identity
- As of
- 2026-09-22
- Revision
- 2026-10-06.1
- Method
- v1.3.0
Definition
The term, in context
AI-assisted draft · approved dataset
Claude Text Watermarking is a content-provenance technique that embeds a statistical signal into text generated by a language model, subtly altering the probability distribution used to select output tokens so the change remains imperceptible to readers yet detectable through targeted analysis. It supports methods for distinguishing machine-generated text from human-authored text without altering visible content, contributing to broader transparency efforts around AI-generated material.
Live readiness status
Status as of the current dateline
AI-assisted assembly · derived results
As of 2026-09-22, verified readiness is 46 (claimed 60, reported 52, gap 14) — Critical evidence strength; current signals suggest Track; not yet.
The readiness fact belongs to the canonical Readiness Verdict for Claude Text Watermarking.
Dataset approval
Human editorial release
Codex — delegated editorial approver (operator authorization 2026-10-06)
- Dataset
- 2026-10-06.1
- Hash
- sha256:0df9a4b64770fab5c2feaca298223eb10279a31b43487556407f496c9b022685
- Approved
- 2026-10-06