Helping Mobile Network Operators Detect Telecom Subscription Fraud and Account Takeovers

Telecom fraud is on the rise globally.

According to the CFCA Telecom Fraud Survey, in 2018 fraud losses as a percent of global telecom revenues grew 37% to $28.3 Billion USD. 

Topping the list of the types of threats facing mobile network operators (MNOs) is telecom subscription fraud, whereby fraudsters use a stolen or synthetic identity to open an account with the goal of obtaining mobile devices and services. With so many operators offering free or subsidized mobile devices, criminals pay little or nothing upfront and can then sell the device(s) or use them to commit other crimes including opening mobile banking accounts and credit cards under real or fake identities. To make matters worse for telcos, these accounts may initially appear as delinquent and remain open for months. 

A less common, but more frightening threat is SIM swap fraud in which a criminal uses stolen credentials to request activation of a new SIM card in order to take over the mobile number. Once the hijacked mobile number is moved to a device that they control, a fraudster can quickly request password resets for sensitive accounts, including mobile banking apps. The losses can be staggering. One Silicon Valley businessman lost one million dollars in a matter of minutes. 

For criminals, MNO telecoms are an easier and lower risk target than financial institutions. 

Fighting Telecom Fraud in the Contact Center

The people who commit fraud in the call center typically do it repeatedly. They commit crime as a profession and become skilled. If they find a weakness, they exploit it as often as possible until they are detected. They want a high return on their criminal effort!  However, their strength is also their weakness, because the skilled fraudster must use the same voice for multiple identities. 

If the fraudster were calling a small company, they likely could not fool the company repeatedly because the company employees would catch on.  But most telcos have thousands of agents; the chance of getting the same agent on different fraud attempts is low! And even then, an agent may not realize the person is the same person the agent spoke to earlier. Therefore, for large telcos, catching fraud happens only after the fact, after the telco has lost money. 

The Power of Voice Biometrics

Voice Biometrics provides a solution to this problem by scanning audio files for the same voice occurring multiple times in the new customer database. By combining the knowledge of the identity associated with the voice, Voice Biometrics enables the telco to identify all new customers with different identities but the same voice within any given period of time, whether a day, week, or month. Because fraud detectors using biometrics run fast enough to process millions of comparisons an hour, it is now possible to stop much of the fraud before fulfilling a new customer order. 

And once the voice is known to be a fraudster, that voice enters a fraud database. All new customer voices thereafter will be compared to the known fraudster voices, and if a match occurs, the telco can immediately flag the new customer order as having a high probability of fraud, stopping the fraud before the fraudulent order progresses further into the fulfillment pipeline. Read about ID R&D’s Telco Fraud Detector to learn more. 

Voice Biometrics can also be used to strengthen existing user authentication — particularly when they are calling to obtain a new SIM card or add services. Voice Biometrics greatly reduces the ability for bad actors to use social engineering tactics to pass verification approaches that rely PINs and passwords, or “secret questions” with answers that are easily researched online.

Fighting Telecom Fraud in Online Enrollment and Identity Verification

More MNO telecom providers are adopting digital onboarding processes similar to what we see in banking for compliance with Know Your Customer (KYC) regulations. Mobile subscribers signing up online are asked to upload a government issued ID and then take a selfie for the two to be matched using face biometrics. 

The problem is that facial recognition cannot determine if a person is real or not and criminals can use high resolution photos and video from the internet to spoof the system.

An AI-driven technology called Facial Liveness Detection distinguishes between a “live” person and a representation of the person — a vital capability for fighting fraud during identity proofing. However, not all solutions work the same way. Some approaches frustrate and confuse users, leading to increased abandonment during digital onboarding. Read about ID R&D’s unique passive facial liveness.

Managing Fraud Effectively

Rather than eliminating all fraud, the goal for MNO telecom fraud prevention should be minimizing exposure. Ineffective fraud management can have a negative impact on the business including:

  • Delayed decision times
  • False positives (or false declines) that result in lost business
  • Overall decline in customer experience
  • High back office costs

All telcos are heavily focused on acquiring new subscribers with easy sign up and great incentives. Can these MNO telecom fraud prevention tools stop all fraud? No, but the right fraud detection tools cost-effectively lower risk and reduce the strain on company resources and margins. 

Contact us if you would like to talk about reducing fraud in your contact center.