Payment fraud

Decide before
you capture.

Send the card payment to the decision endpoint; the engine tells you to continue, step up to 3D Secure, review or decline.

Behavioural signals

  • copyPasteCount2
  • tabSwitchCount5
  • typingCadenceScore0.21
Decisionforce_3d

Send a token or fingerprint, never the card number.

What's included

The known traces of card payment fraud.

  • 3D Secure step-up

    For risky transactions the decision says to step up to 3D Secure; it never passes without 3D.

  • Card testing

    Ready templates for card testing patterns.

  • Merchant risk templates

    Ready templates for merchant-side risk patterns.

  • Behavioural signals

    Paste count, tab switches, mouse movement and typing cadence travel with the decision call.

  • Velocity rules

    Once you feed transaction events, history-based rules such as transactions in the last 24 hours work.

  • Country and instalments

    Card country mismatch, amount thresholds and instalment anomalies.

How it works
  1. 01

    Feed events

    Send transaction events to the event backbone.

  2. 02

    Ask for the decision

    Send the transaction to the decision endpoint before paying.

  3. 03

    Apply it

    Continue or step up to 3D based on the decision.

  4. 04

    Monitor

    Read decisions from the decision logs.

Make fraud decisions
with your own rules.

Let's set up your flows and thresholds together with our team.

For enterprise inquiries: [email protected]

Payment fraud — Fraud and Risk | Lunixi