What Does 'C2PA Manifest Detected' Mean?
Published August 4, 2026
When you run a file through our C2PA Checker, you'll see one of a few distinct outcomes. Each means something specific and limited — here's what each one does and doesn't tell you.
"No Content Credentials detected"
The tool found no C2PA manifest in the file. This is the most common result, for one of several reasons:
- The file was never produced by C2PA-aware software
- A C2PA manifest existed but was stripped — by a platform on upload, by re-saving the file, by a screenshot, or by a metadata-cleaning tool like our own Metadata Cleaner
- The file predates the software or camera adding C2PA support
This result says nothing about whether the content is authentic, AI-generated, or edited. It only means: no provenance claim survived to this point.
"Content Credentials detected"
The tool found and parsed a manifest. You'll then see a validation state, which is the part people most often misread:
- Trusted — the manifest's cryptographic signature is intact, and it was signed by an issuer whose certificate is recognized by the trust list the checker uses.
- Valid — the signature is intact and internally consistent, but the signer isn't on a recognized trust list (or trust couldn't be evaluated). The data hasn't been tampered with since signing; who signed it just isn't independently vouched for by this checker.
- Invalid — something about the manifest failed validation: a broken signature, structural corruption, or a mismatch between the manifest and the actual asset bytes. This can happen legitimately (the file was edited by software that touched the pixels without updating the manifest) as well as maliciously.
None of these three states are a verdict on whether the content itself is real, AI-generated, or trustworthy — they only describe whether the manifest's own signature checks out. See How C2PA Content Credentials Work for why that distinction matters.
The manifest details
When a manifest is detected, our checker also shows, where available: the claim generator (the software that made the claim), signer/issuer information, assertions (the specific claims made), and ingredients (source files referenced in the provenance chain). Read these as a disclosure from whoever created the file — informative, but not independently verified beyond the signature check itself.
"Unable to analyze"
The file format isn't one the checker's WASM-based reader supports, or the analysis failed for a technical reason (corrupted file, unusually large asset). This is different from "not detected" — it means the tool couldn't reach a conclusion either way, not that it concluded there's no manifest.