Frequently Asked Questions

Sourcemark is consent infrastructure for AI training data. It lets creators declare how their work can and cannot be used for AI training, and gives AI companies a machine-readable registry they can check before ingestion.

Sourcemark is built for creators, rights-holding organisations and AI companies. Creators can record their AI training choice, organisations can declare across catalogues and AI companies can check the registry before training.

No. Sourcemark is not built to block AI. It is built to make creator choice visible, whether that means work is not available for AI training or available by licence. AI companies are part of the system too.

Robots.txt is a request to web crawlers. It was not designed for AI training consent. It has no audit trail and it is often ignored. Sourcemark creates timestamped, machine-readable records that can be checked and verified.

No. Sourcemark records consent and provides a contact pathway for licensing. It does not set prices, negotiate deals or process payments.

No. Sourcemark records who made a declaration about a specific file and when. It does not certify legal ownership or verify authority.

No. Unknown means there is no registered Sourcemark signal for that work. The absence of a declaration is not consent.

Sourcemark currently supports images, PDF documents and video.