Insurance claim photo fraud detection
A motor or property claim now starts with photos taken on a phone, and an image app can add a dent, a crack or a broken light to a real photo in under a minute. TrueSecurix checks each claim photo, estimate and bill before the payout, and tells your claims team whether to pass, review or flag it, with the evidence.
In our held-out test of 31,966 images, 0 of 2,526 genuine business photos (shops, markets, receipts, documents) were flagged and 87.8% of manipulated images were caught. Read how we measured
What claims and fraud teams are seeing in claim files now.
What claim fraud looks like today
Damage that was never there
An AI editor adds a dent, a scratch or a shattered light to a photo of a vehicle that was never damaged.
An accident that never happened
A whole scene made by an image model: the car, the road and the damage, with no camera involved.
Pasted and cloned damage
Damage copied from another photo, or from another part of the same photo, and pasted in by hand.
Estimates and bills with new amounts
A repair estimate, a garage invoice or a hospital bill photographed or scanned, then edited to raise the amount.
Image and identity checks run on the same account and the same credits: one credit per check.
Which checks apply, and what you get back
Each image goes to POST /v1/image/authenticity; selfies, IDs and statements go to POST /v1/verify. Every response carries an outcome and the evidence behind it, and every check is written to your history.
| What you receive | Check | What the response gives |
|---|---|---|
| Vehicle and property damage photos | Image authenticity | Real, AI generated, AI edited or digitally edited, with the indicators and the located edited areas. |
| Repair estimates and garage invoices, as photos or scans | Image authenticity | The same four answers, with the edited areas located so a surveyor knows where to look. |
| Medical bills and pharmacy receipts, as photos or scans | Image authenticity | The same four answers and the evidence behind them. |
| Claimant selfie and ID, where your process collects them | Identity checks | Deepfake selfie detection, face match to the ID photo, and Aadhaar, PAN and passport checksum validation. |
Nobody is rejected on a guess. The model tells you which files a person should look at.
Three outcomes your claims team can act on
The decision field turns the answer into an action. Route on it directly in your claims system.
Accept it
No sign of generation or editing. The claim continues through your normal assessment.
A person decides
Some evidence points to a manipulation, not enough to accuse a genuine claimant. A claims handler takes a short look, guided by the class it leans toward.
Treat as manipulated
Strong evidence of generation or editing, with the indicators and the edited areas marked. Hold the payout and ask for a survey or the original files.
How a team rolls it out
Most of the work is wiring three outcomes into the claims system you already run. You can build all of it before you spend a credit.
One request, one image, one answer
curl https://truesecurix.com/v1/image/authenticity \ -H "X-API-Key: $TRUESECURIX_API_KEY" \ -F "image=@rear_bumper.jpg" \ -F "reference=claim-30917"
Field by field in the API reference.
- Build against a test key. A test key returns a simulated answer at no charge, including the exact review shape when you send
reference=review, so your claims system handles all three outcomes before any live image is sent. - Run a proof of concept on your own labelled images. A folder of real and manipulated claim photos, estimates and bills from your own files, with a CSV of labels, goes through the batch script. You get accuracy per class, the confusion matrix, genuine images wrongly flagged and the share sent to review, on your data.
- Send results into your claims system. Signed webhooks post
check.flagged,check.reviewandcheck.completedto your HTTPS endpoint, signed with HMAC-SHA256 and retried for about a day. - Route reviews to a person. A review carries the class the evidence leans toward and the indicators behind it, so a reviewer knows what to look for. Every check stays in your history with its model version, and exports to CSV.
Measured on images it never saw
The image model was measured on business photos of shops, markets, receipts and documents, plus general photos, in a held-out test of 31,966 images never used in training, threshold setting or model choice.
Your own images are the measure that counts for your use. A proof of concept on your own labelled images gives you the same figures on your data, with the per-image CSV behind every number.
Questions risk, operations and engineering teams ask first.
Questions, answered
Can it tell real damage from damage added by an AI editor?
It answers AI edited when part of a real photo was redrawn by an AI editor, and marks the areas where it found the change. When the evidence is not strong enough to accuse a genuine claimant, the photo goes to review with the class it leans toward instead of being flagged.
Does it reject claims?
No. Each photo comes back as pass, review or flag with the evidence. Your claims process decides what happens next, and a review always goes to a person.
Can it check estimates and bills sent as PDF files?
The image check takes JPEG, PNG and WebP images up to 15 MB. Send a photo or a scan of the estimate or bill, or convert the page to an image first.
How accurate is it on claim photos?
The model was measured on business photos of shops, markets, receipts and documents, plus general photos, in a held-out test of 31,966 images. It has not been measured on claim photos specifically. A proof of concept on your own labelled claim photos is how you measure it for your use.
Check claim photos before the payout
Try a photo in the live demo now, or create an account for 100 free checks. No card is needed, and test keys are free while you build.
Running lakhs of checks a month? See enterprise terms and a proof of concept.