Overview

FraudStop was created as a response to the increasingly sophisticated fraud techniques we were seeing in the DCB/VAS market. Where previously we had seen misleading advertising flows and prefilled SMS messages tricking consumers into making payments through social engineering techniques, we started to see MSISDN injection and mobile app based malware triggering transactions without the consumer’s knowledge through the use of technology.

FraudStop is a cyber-security platform that identifies fraudulent payments in the carrier billing environment.

FraudStop differentiates between genuine transactions by users and fraudulent transactions generated on users’ handsets without their knowledge or injected into the network layer with a range of measures including: IP validation, Device fingerprinting, Click behaviour, Browser/App presentation, i-Framing, Suspicious traffic sources
Each transaction is validated against 100s of criteria within milliseconds to identify fraud, including: App malware, I-Framing, Bots, URL spoofing, MSISDN Injection

A token is assigned to a user with suspicious parameters recorded against the token. Transactions without a valid token, or with any of Fraudstop’s blocking criteria flagged, are notified to the payment provider.

To counter new and emerging threats, FraudStop uitlises a detection engine using machine learning to continuously evolve and learn. This is supported by data from its sister platform, FraudScan, which identifies rogue applications and new suspicious traffic sources by automated testing, as well as manual testing.

FraudStop is deployed via a simple script, typically on the payment page. The solution provides authorised parties with real-time access to track attempted fraud, which is delivered through a dashboard.

FraudStop runs on AWS infrastructure hosted on servers in London, Sao Paulo and Singapore to minimise latency. We also operate a fleet of locally hosted mobile devices across the 30+ countries Fraudstop operate.