Know which photos were generated or edited by AI.
Reads Content Credentials (C2PA), XMP and generator metadata to badge AI-generated and AI-edited media in the grid, the viewer and search filters.
AI images arrive in galleries the same way any other image does: shared, saved, screenshotted. Pixel Gallery reads the provenance data that generators and editors write into files and labels what it finds.
What is read
- Content Credentials (C2PA): the industry manifest embedded by supporting cameras, editors and generators.
- XMP and EXIF fields: generator names and edit histories written by tools.
- PNG chunks: generation parameters saved by many image generators.
From these, the app distinguishes AI-generated from AI-edited media and records the named source where one is declared.
Where it shows
- A badge on the grid tile.
- A row in the viewer’s info sheet with the declared source.
- The “AI content” quick filter in search, so you can find or exclude AI media at once.
Limits, stated plainly
This is metadata detection. A screenshot of an AI image has no provenance data and will not be flagged. Manifest signatures are not verified in this version, so a declared source is a claim, not a proof. The app says what the file says, no more.
Related
Detection runs as part of the on-device pipeline and never sends the image anywhere.
Questions about ai content detection
Does it analyze the pixels?
No. Pixel-level detectors are unreliable, so the app reads provenance metadata instead. Files whose metadata has been stripped are not flagged.
Are Content Credentials verified?
The app reads the credentials and reports the declared source. The certificate chain is not cryptographically verified in the current version.
Free on Google Play
Requires Android 13 or newer. No account. Optional one-time Premium upgrade.