Jaskaran Singh, Senior VP for Big Data & Analytics at Mobileum
Every few months a blaring headline pops up alarming us about hacking or some other security or privacy failure with a particular mobile network, which over time has put major dents in the credibility of the industry. In fact, telecoms is in a pitched battle over the security issue, trying to stem global fraud losses estimated to be bleeding the industry of $29.2 billion annually, roughly 1.27 percent of global telecom revenues.
In terms of hitting the operator’s bottom line the hardest, Interconnect bypass (for example SIM Box) events are the most common problem, accounting for $4.27 billion, while International Revenue Share Fraud (IRSF) is the costliest issue, adding up to $6.10 billion in losses, according to the Communications Fraud Control Association (CFCA). The question is, how did security lapses get so bad, and so costly for operators?
The problem is two-fold. First, fraudsters are getting increasingly sophisticated in their techniques. They are experts at avoiding detection, quickly finding new ways around the latest security measures. And secondly, the legacy systems currently in use by many of today’s operators just can’t keep up.
Here’s the root problem with legacy, rules-based, systems; while they’re good at spotting and reacting to known issues, they struggle mightily at picking up emerging threats, which is critical to preventing fraud in the long-term. Also, they’re limited by finite databases which, at best, can hold up to three months’ worth of data – that’s no use for tracking discrete patterns of potentially fraudulent activity over the long haul. And scratch the surface a little deeper, and you’ll find even more blind spots in traditional fraud detection methods, such as high rates of false positives and delayed response times after a fraud already occurred.
That’s where big data analytics, using AI, neural networks and machine learning, comes in, to help shed light on those glaring blind spots. High capacity, intelligent systems are quick to learn what constitutes potentially suspicious activity. And by performing very complex analyses, the technology can pick up the subtlest patterns in a wide range of data - to a degree once seen as impossible.
Unfortunately, there’s a general misconception that relying on machine intelligence can be dangerous for operators, but that’s not the case. AI’s role is to detect the previously undetectable and do it quickly, but operators still have the power to set strict controls and train systems in what to look for, what level of sensitivity is appropriate, and what action to take.
In the end, it’s about making fraud detection more intelligent, targeted, accurate and cost-effective. Advances in AI are not only improving hit rates and speed of discovery, but their saving fraud analyst teams from sifting through reams of false alerts, driving up cost-efficiency.
A carefully implemented AI solution that is continuously adapting, and aids human decision-making, offer the ability to level the fight against fraudsters. However, it will take a bit of rebellion to break away from existing approaches, as well as guidance from bold solutions providers to help operators see it through.
In part three of this blog series we’ll discuss real-world examples of how Mobileum sees fraud detection in a new light, putting operators on the offensive - so stay tuned!