SCBX
Weekly Tech
Anthropic Begins Watermarking Claude-Generated Text
Claude models launched on or after August 2, 2026 embed machine-readable marks at generation time: an imperceptible statistical watermark in generated text, and signed provenance metadata on generated files. Coverage spans the API, Claude, Claude Code, Claude Cowork, and Claude Tag, applied worldwide under Anthropic's commitments to the EU AI Act's Article 50(2) Code of Practice. Anthropic notes a detected mark shows only that content passed through Claude — proofreading or translation of a user's own writing carries it too — and that no mark does not imply human authorship, since heavy editing can destroy the signal.
Connect with us
via LINE OA.
Stay updated with R&D. Scan the QR.

