Stopping Losses from Voice Fraud: Artificial Intelligence or Global Number Intelligence?

Voice Fraud

Fighting voice fraud is more important for carriers than it’s ever been. Carriers have to explore new ways to combat fraud and take a holistic approach to identifying and mitigating it across their networks. 

According to the Global Leadership Forum’s (GLF) Fraud Report 2020 “Tackling Fraud: A Collective Industry Effort”, 72% of carriers reported fraud to be a “top” or “strategic” priority. Almost 60% of carriers are expected to invest more in fraud monitoring/prevention infrastructure in the next 12 months and 39% say they will grow their fraud teams over the same time period.

Increasingly sophisticated fraudsters, downward pressure on margins and a challenging business environment where roaming revenues have disappeared, are all driving the need for effective fraud management systems. The overall health of the voice ecosystem depends on maintaining margins and ensuring losses to fraud are minimised. Otherwise, carriers will be winning business while margins are eroded, not just by OTT competition, but by organised crime.  

There is no one silver bullet, but there are opportunities to expand capabilities and support systems with new intelligence. No single platform or technique will stop all voice fraud, but carriers can increase their chances of success by looking at the problem from multiple angles.

As criminals are agile innovators and methods of fraud are continually changing, carriers have to take a holistic approach to fraud management and explore new ways to stay ahead of the game.     

In an ideal scenario, a carrier would combine Bayesian rules and Machine Learning along with human supervision and intervention.  However, rules and patterns based on historical data analytics aren’t always best suited to spot fraud in the future, hence the importance of some logic founded on the use of the most fundamental aspects of calling – phone numbers and call path.  It is about combining complex technologies and systems to examine the call while it is occurring, with deeper levels of numbering intelligence to identify high-risk B numbers.

All approaches need to ensure that fraud is identified as it is attempted rather than after the fact. By that time, it is too late and losses have already occurred. It can take anywhere from 30 days to six months for any dispute with providers to conclude.  

While most carriers have invested in fraud management systems to examine fraud directly in the call path and deployed rule-based policies, they often overlook the value that global numbering intelligence can deliver. Global number intelligence can support existing systems and help to tackle the most prevalent type of voice fraud – International Revenue Share Fraud (IRSF). The GLF noted that 45% of carriers reported increases in fraudulent traffic and impact vs. 2019, with IRSF and Wangiri increasing most significantly.

In the case of IRSF, carriers that rely only on one technique alone often deliver false positives and block legitimate traffic. This causes unnecessary losses in revenue and can be improved through identifying high-risk numbers. IRSF is commonly found in unallocated number ranges, international premium rate numbers (IPRN), premium rate numbers (PRN), leased ranges, and non-standard ranges.

Carriers can add global number intelligence to their anti-fraud strategies and increase the precision of their traffic blocking, while ensuring they maximise their revenue. They can analyse phone numbers down to the last 2, 3 and 4 digits.

While many carriers have internal data sets, they can be immediately expanded to increase granularity of data across more markets and operators. XConnect provides carriers with access to more than 6 million records of global numbering range data, and number portability data for voice routing spanning more than 35 countries.

The outcomes for carriers include the ability to analyse A and B-numbers in real-time and stop false positives. Global number intelligence is added to their overall approach to fraud and they gain an answer to growing IRSF losses. New data sets augment and support existing fraud management systems, while giving carriers increased visibility into the traffic that they are carrying.

Ultimately, carriers that take a holistic approach to fighting voice fraud gain greater control over their traffic, benefit from predictable margins and avoid lost revenue opportunities.

Latest Posts