Skip to content

Device Intelligence & Fingerprinting

Definition

Device intelligence uses signals from the user's device — hardware attributes, software configuration, and behavioral patterns — to assess fraud risk before, during, and after eKYC verification.


Device Signals

Signal What It Reveals Risk Indicator
Device fingerprint Unique device identifier Multiple accounts from same device
Root/jailbreak Device security compromised Can inject camera, modify app
Emulator Not a real phone Likely automated attack
Virtual camera Fake camera feed Injection attack
GPS spoofing Fake location Location fraud
VPN/proxy Hidden real IP Masking true location
Device age How long device has been seen New device = higher risk
App tampering Modified app binary Bypass security checks
Screen recording Screen capture active Data exfiltration risk
Debug mode Debugger attached Analysis/attack in progress

Device Fingerprinting Techniques

Technique Persistence Accuracy
Hardware ID (IMEI, serial) Permanent High but privacy-restricted
Canvas fingerprint Semi-persistent (web) Medium
WebGL fingerprint Semi-persistent (web) Medium
Audio fingerprint Semi-persistent Medium
SDK-generated ID Per-install High within app
Behavioral fingerprint Session-level Medium

Providers

Provider Key Feature
SEON Device fingerprinting + email/phone intelligence
ThreatMetrix (LexisNexis) Largest device identity network (6B+ devices)
BioCatch Behavioral biometrics + device
Sardine Device + behavior + transaction
Castle Account security + device

Key Takeaways

Summary

  • Device intelligence is a critical pre-verification signal — detect fraud before expensive processing
  • Same device + multiple identities is the strongest fraud ring indicator
  • Root/emulator/virtual camera detection prevents injection attacks
  • Device fingerprinting faces privacy restrictions (GDPR, App Tracking Transparency) — balance needed
  • Combine device signals with verification + behavioral signals for comprehensive risk scoring