Use this sequence when an image’s origin matters. It gathers separate clues without asking one test to prove too much.
1. Inspect the image
Look at the scene as a whole, then check text, geometry, reflections, object connections, and repeated textures. A visual oddity can come from photography or editing. Treat it as a question to investigate.
2. Run the AI Analyzer
Upload the image to the AI Analyzer. Record the score and verdict, then read the limitation note. The model estimates class resemblance. It does not prove the image is real, generated, or made by a specific tool.
3. Inspect EXIF fields
Use the EXIF Detector to check available camera, date, lens, software, orientation, and GPS-presence fields. Metadata describes values in this copy of the file. It may be absent or edited. Motchilllz does not display exact GPS coordinates.
4. Check software and provenance
A software field can name an application that wrote the file. It does not identify every change or prove manipulation. If the file includes Content Credentials, use a compatible verifier to check the signed record and interpret the claims. Motchilllz does not currently validate C2PA.
5. Find the source
Look for an original post, photographer, publication, or earlier version. Compare context and captions as well as pixels. A reverse image search may find matching copies, but matches do not by themselves establish which copy came first.
6. Keep uncertainty in the account
Describe what you observed: a classifier score, a metadata field, or a source record. If those clues disagree or the original is unavailable, say that the origin is unresolved. See how to tell if an image is AI-generated for more detail on individual checks.