General FAQs
- Where can I find signed sample assets?
- How can I get started with Content Credentials on Android or iOS?
- Does a Valid validation state mean that the signer is trusted?
- Can a manifest carry custom sensor or verification information?
- Where are embedding rules for TIFF-based and base-media file formats documented?
- Where can I report a CAI security concern?
- Why might an HEIC thumbnail not appear in Verify?
Where can I find signed sample assets?
Use Proofmode's public
Baseline repository has signed sample
assets, but be aware that much of its content uses older C2PA 1.x support. Use the contentauth/example-assets repository to submit signed assets for testing. Neither repository guarantees a
complete, current conformance test suite.
For location metadata, see the Proofmode
Android 3.0.2 release candidate
and its use of c2pa.metadata to include geolocation as a created assertion.
Older conformance samples used stds.exif, described as deprecated but
still valid. See the
Kotlin metadata-generation code.
Sources:
- Baseline and older sample versions; April 3, 2026
- Android geolocation implementation; April 3, 2026
- Older sample assertion format; April 3, 2026
- Metadata-generation code; April 3, 2026
- Shared signed-asset repository; April 3, 2026
How can I get started with Content Credentials on Android or iOS?
Use the mobile SDKs for Android and iOS.
Does a Valid validation state mean that the signer is trusted?
No. There are three manifest states: well-formed, valid, and trusted. Trusted manifests are also valid, and valid manifests are also well-formed. The reverse does not follow.
A manifest signed with a test certificate can have a Valid state while
validation_results.activeManifest.failure contains
signingCredential.untrusted. This follows the manifest-state model.
Do not label a manifest trusted merely because its state is Valid.
Inspect the trust-related validation results, and distinguish trust that
was not evaluated from a failed trust check. An issuer name alone does
not establish trust.
See the manifest states specification.
Sources:
- Valid-but-untrusted result; July 29, 2026
- State hierarchy; July 30, 2026
- Specification reference; July 30, 2026
Can a manifest carry custom sensor or verification information?
Yes. You can add custom assertions to a C2PA manifest. A validator that understands that assertion could use its information to identify where changes occurred.
That is an extensibility mechanism, not endorsement of the proposal's claimed security guarantees or a promise that existing validators understand it. A camera/editor workflow needs compatible implementations; custom assertions alone do not provide that coordination.
Source: custom assertions and validator support, April 20, 2026
Where are embedding rules for TIFF-based and base-media file formats documented?
See the Embedded Formats annex of the Content Credentials specification. for embedding rules relevant to TIFF-based Sony formats and the base-media-based Canon CR3 format.
This is a specification pointer, not confirmation that every raw camera format is supported by a particular SDK or that camera samples are available.
Source: Embedded Formats annex, August 17, 2026
Where can I report a CAI security concern?
Use the CAI contact form to report a security concern. Do not assume a dedicated security inbox, a response deadline, or public status updates are available.
Sources:
Why might an HEIC thumbnail not appear in Verify?
Signing and display are separate issues. If a thumbnail's format changes to
application/octet-stream during signing, use a v2 claim: set
claim_version to 2, or omit it if your version defaults to v2.
Display also depends on browser support. As of September 2026,
image/heic thumbnails displayed in macOS Safari but not Chrome.
A correctly embedded thumbnail therefore did not guarantee display in
every browser. These observations are historical, not a current browser
compatibility matrix.
Sources: