Claude AI Content Now Marked With Invisible Watermarks Globally

The watermark is embedded at the model level, so it applies across all access points, including Claude Platform, Claude Code, Claude Cowork, and Claude Tag. Anthropic said the marking will be present globally wherever supported Claude models are available, not just in Europe.

For files such as SVG, PNG, and JPG images, Claude will attach digitally signed provenance metadata based on the Coalition for Content Provenance and Authenticity (C2PA) standard. This metadata can indicate that a file was processed by Claude and help determine if it was later altered, though support may vary by file type and platform.

The watermark is designed to be invisible and not affect readability or quality. It can survive copy-pasting and some editing, but heavy editing, paraphrasing, translation, or combining text with other material could make detection harder. For files, format conversion, screenshots, or re-saving can strip provenance information.

Anthropic is developing tools for users and third parties to check for Claude markings and plans to publish technical documentation on detection. However, the company stressed that a detected mark is not definitive proof of authorship, since someone could use Claude to translate or proofread their own work, and the absence of a mark does not prove human authorship.

Models released before Aug. 2 are covered by a transition period under EU rules, and Anthropic is working to add marking capabilities to them. The company also said it will provide further technical guidance for developers building with Claude, who must assess their own obligations under the EU AI Act.