ChronoVerify

Insurance claims

Claims teams receive photos of damage from strangers, at scale, with money attached. ChronoVerify triages each image's capture time and provenance so adjusters spend their attention on the ones that need it.

The workflow

  1. Verify on intake. Call the API as each claim photo is uploaded, by URL or file.
  2. Read capture time and device. Confirm the photo was taken around the loss date on a real camera, not pulled from the web.
  3. Flag for review. Route metadata_anomaly and manipulation_indicated to a human; let consistent results flow.

Which verdicts matter

A consistent result with a plausible capture time and device supports the claim narrative. manipulation_indicated or metadata_anomaly is a triage flag for an adjuster, never an automatic denial.

What this can and cannot tell you

ChronoVerify cannot prove the depicted damage is real or recent; it reports what the file's own evidence says. A re-saved or screenshotted photo often returns inconclusive, which is not evidence of fraud. Never deny a claim on a verdict alone.

One call

curl -X POST https://chronoverify.com/v1/verify \
  -H "Authorization: Bearer cv_live_..." \
  -F "url=https://claims.example.com/photo.jpg"

Omit the Authorization header to use the free, rate-limited public path. Full field reference is on the method and API page.

Common questions

Can this detect a staged or AI-generated claim photo?

No. ChronoVerify is provenance-first, not a deepfake or AI-generation detector. It flags missing or contradictory provenance for human review; it does not decide whether a scene was staged.

What should auto-pass versus go to review?

A common pattern is to let consistent results with a plausible capture time flow, and route metadata_anomaly and manipulation_indicated to an adjuster. Set the confidence threshold to your risk tolerance.

See a verdict on a real photo.

Try the free verifier