Telecommunications firms are grappling with AI for fraud detection and prevention.
Virgin Media O2’s AI-generated grandmother, Daisy, keeps fraudsters engaged—effectively turning the tables on scammers by using their tactics against them.
Telecom fraud causes not only financial losses but also erodes customer trust.
AI models now analyze billions of call detail records (CDRs) daily to detect complex patterns of fraud like SIM box fraud.
Device fraud revolves around purchasing high-end devices on installment plans and paying only one installment before disappearing.
Wangiri scams are another problem. Here, victims receive missed calls from premium numbers and are charged exorbitant fees when they return the call.
Legacy systems complement AI-driven solutions by providing foundational insights.
A hybrid approach that combines the stability of traditional systems with the adaptability of AI models is often preferred to ensure an effective way to address both traditional and emerging threats.
Fraud detection requires frequent model retraining.
The telecom industry loses $40 billion annually to fraud globally.