Read the file underneath the picture
A scan inspects supported structures in the selected file. Where readable, it extracts source declarations and personal fields. This is metadata inspection, not a visual judgment about whether an image looks AI-generated.
Separate AI declarations from personal information
Content Credentials and digital source types can describe how media was created or edited. Camera, location and creator fields serve other purposes. The scan distinguishes these findings so you can judge the value of cleaning your sharing copy.
Confirm access before cleaning
You can view a scan without an account. To clean, register or sign in and activate unlimited access through checkout. No successful-clean result is shown merely because an account was created or a payment page was opened.
Remove the supported records
The cleaner processes supported C2PA, XMP, EXIF and IPTC metadata while leaving encoded media data unrecompressed. Because those blocks can include useful display or attribution information, retain the original and inspect your saved output.
Verify the new file
The output receives its own scan. Verification concerns supported metadata targeted by the cleaner; it does not authenticate a scene or check a social network’s private labeling systems. Incomplete parsing prevents a verified download.
Videos need their own final check
A video assembled from clean source images is a new export. Rendering can introduce new metadata, so inspect the finished MP4 or MOV rather than relying on a source frame’s result.
Published labels are separate
Cleanup prepares media before upload. It cannot alter a live post, guarantee reach or remove every reason a platform may display AI information. Use the appropriate platform guide to understand that distinction.