Provenance guide

Content Credentials Explained

Short answer: Content Credentials are cryptographically bound provenance records defined by C2PA. A validator can check whether their assertions remain bound to the inspected asset and whether the signature is valid. Trust still depends on the signer and the claim; credentials do not decide whether depicted content is true or good.

Presence, validity and trust are different states

StateMeaningSafe interpretation
Not foundNo embedded or discoverable manifest was available to the checker.Neutral. Export, screenshots or platform processing may remove credentials.
Marker foundThe file structure indicates a possible C2PA manifest.Do not rely on it until a cryptographic validator reads it.
InvalidThe manifest, signature or content binding did not validate.Treat the credential as unusable; investigate the failure code.
Valid, signer untrustedThe data validates, but the signer is outside the checker’s trust list.The binding may be intact; signer identity and authority still need review.
Valid and trustedThe binding validates and the signer meets the configured trust policy.Rely only on the specific signed assertions, not on unstated conclusions.

What a trusted credential can support

A credential can record claims about creation, edits, ingredients, tools and digital source type. Validation can show that those signed assertions are still associated with the inspected asset and have not been altered outside the recorded chain. The C2PA specification is the primary technical source.

What it cannot prove

C2PA explicitly separates verifiable provenance from value judgments. A sound workflow reads the assertions, validates the binding, evaluates signer trust and then combines that evidence with source context.