Nine checks per document.
Each check looks at something different and returns its own finding. Together they produce an overall grade from A (unremarkable) to F (highly suspicious) — traceable, because you see every individual finding.
How the check works.
The authenticity check runs alongside the extraction — same step, no second upload. That way you can spot edited or AI-generated vehicle registrations before their data reaches your system. Enable the check per document or permanently in your account settings.
Switch the check on
You use the same flow as for extraction and enable the authenticity check on top — in the dashboard or per request.
Nine checks at once
Image details, editing traces, preview image, capture pattern, copied areas, security pattern, AI detection, content contradictions and duplicates all run independently.
Overall grade A to F
The individual findings are combined into one grade — deliberately conservative, so genuine documents don't get flagged needlessly.
A finding per check
You get the extracted data as usual — plus the risk assessment with a grade and a reason for every check.
What the result looks like.
The extracted data stays as it is. The risk assessment is added on top — only when you switch it on, never automatically.
- Image details: traces of an image editor
- Editing: strong anomaly around the VIN
- Capture pattern: deviation around the owner field
- Copied areas: none found
- Duplicate: document number already submitted with a different VIN