Okta
flow-image

How Okta uses machine learning to automatically detect and mitigate toll fraud

Published by Okta

The Okta White Paper explains how machine learning (ML) is used to detect and mitigate toll fraud, specifically International Revenue Share Fraud (IRSF), which exploits SMS and voice-based MFA for financial gain. Fraudsters generate high-cost international calls or texts during authentication flows, leading to significant financial losses for businesses. Okta’s anti-toll fraud system combines heuristic detection, unsupervised ML models (isolation forest algorithm), and risk-aware rate limits to analyze transactions in real time. This approach improved detection by 20% without increasing false positives, protecting authentication systems while ensuring reliable user access and minimal disruption.

 

 

Download Now

box-icon-download

Required fields*

Please agree to the conditions

By requesting this resource you agree to our terms of use. All data is protected by our Privacy Notice. If you have any further questions please email dataprotection@headleymedia.com.

Related Categories Security, Site Security, Artificial Intelligence, Deep Learning, Unsupervised Learning, Reinforcement Learning, ML Algorithms, Data Preprocessing, Model Training, Inventory Management, Predictive Maintenance, Logistics Optimization, Quality Control

More resources from Okta