Skip to main content
Add automated checks that flag suspicious behavior and stop high-risk users, before they complete onboarding. You can configure multiple fraud signals, each with custom logic and fallback actions, to suit your risk appetite. Fraud checks run as part of your verification flow when enabled during flow configuration.

​
How It Works

Each fraud signal works as a rule engine. You set a condition. If the condition is met, Dojah automatically flags the verification and applies a status (like pending, failed, etc.). You can configure multiple rules. Here’s what’s available:

​
Available Fraud Checks

Match fields like name or date of birth against government records or internal databases. If selected fields don’t match, you can flag the user or set their status to pending.Use this when you want to:
  • Enforce strict identity consistency across sources
  • Auto-pend or block unverifiable identities
Set a minimum age for users. If the age derived from the user’s submission falls below the threshold, they are flagged automatically.Ideal for:
  • Age-restricted products (e.g., credit, crypto, alcohol)
  • Compliance with local age requirements
Add an extra check that confirms the selfie is live and not a spoof. Combine this with a fallback status to catch suspicious or failed attempts.Best used when:
  • You want to prevent static image or replay attacks
  • You need to enforce real-time biometric presence
Screen users against watchlists such as:
  • PEP (Politically Exposed Persons)
  • Sanctions
  • Warnings
  • Adverse Media
You can configure the AML confidence threshold and set outcomes based on match results. Ideal for:
  • Regulated products (banks, fintechs, crypto)
  • High-risk user monitoring
Automatically detect if the same ID or user details (email, phone, etc.) have been used in previous verifications.Use cases:
  • Preventing synthetic identity abuse
  • Enforcing one-user-one-account policies
Check whether a user’s IP or device is flagged in Dojah’s internal blacklist.This helps:
  • Catch repeat offenders using same infrastructure
  • Flag suspicious devices accessing from high-risk IPs

​
Setting Outcomes

Each fraud rule lets you set an outcome when a match, mismatch, or threshold is triggered. You can choose from:
  • Pending - Review required
  • Failed - Terminate onboarding
  • Success - Allow to proceed
This gives you granular control over how fraud signals affect onboarding decisions.
⌘I