In the past telcos have been forced to perform expensive and piecemeal tests of inbound voice calls, or else they must simply trust that the origin of each call they receive is consistent with the calling line identification (CLI) presented to them. This has always left the telcos and their customers victim to a wide variety of frauds. Now a simple out-of-band 'handshake' confirmation message is being passed between the originating and terminating call providers at the beginning and end of each call, giving both parties absolute confidence that no intermediary has interfered with its routing or duration. The power and precision of this fraud prevention technique means it is no surprise that the community of telcos that use the AB Handshake is rapidly growing in size.
The telecoms industry has long known that most fraud occurs because of bad actors working amongst the intermediate carriers that transit calls between their origin and destination. These carriers may hijack a call, stop it short of its destination, stretch its duration, or use simboxes and refiling to cheat on what they pay to honest telcos. These tricks and scams succeed because the originating and terminating network have previously had no effective way of working together to collectively protect themselves and their customers. This weakness has been overcome by the AB Handshake, which uses a secure communications protocol between servers belonging to the originating and terminating operators to confirm they have a consistent understanding of every call that passes between their networks.
The key parties to any call are the customer that dials a number and the customer who picks up at the other end, as well as the networks they each choose to use. The dialling customer effectively pays all the businesses in the end-to-end chain, whilst the customer who accepts the call signals that a revenue-generating transaction may commence. However, the previous absence of any independent verification between the originating and terminating networks means fraudsters repeatedly spoof end users or trick real end users into doing things they would not otherwise do.
Many international carriers have signed the anti-fraud code of conduct written by the ITW Global Leaders' Forum (GLF). However, any business person knows that having a written agreement is not the same as being able to show that the agreement is followed in practice. Principle 3 in the GLF code says:
"Destinations
Identified fraudulent number ranges and destinations to be blocked
Carriers will actively monitor their individual traffic patterns to identify fraudulent number ranges and destinations, and take appropriate measures individually to block fraudulent traffic as soon as technically feasible."
The weakness in this statement is that it only expects carriers to work as individuals, and never as a collective. If taken at face value, this principle says carriers have the right to interfere with revenue-generating transactions by deciding which calls are or are not fraudulent. The result is a lot of argument between businesses about when calls should be blocked and little confidence that fraud is being tackled consistently or effectively. The old fraud management paradigm of every business protecting their own interests is failing to provide the protection that the telecoms industry needs as a whole.
The telecoms industry has spent decades investing in statistical methods of identifying fraud. This began with simple automated rules relating to high usage and unorthodox traffic patterns. Over time the rules became more complex. Now we have begun to use machine learning to identify anomalies indicative of fraud. The problem is that performing a statistical analysis is not the same as knowing for a fact whether a call is fraudulent or not. The most sophisticated fraud management systems will continue to suffer false positives and false negatives so long as they rely on statistical analysis of data.
It is bad to allow fraudulent calls to be connected. It is worse that legitimate revenue-generating calls are sometimes blocked. There have been cases where an entire country has been unable to receive calls from some European locations because fraudsters have abused number ranges and because of the limitations of the statistical approach to identifying fraud. The AB Handshake community is breaking the cycle of failure by adopting a completely new paradigm to fraud prevention. Instead of trusting intermediaries to identify fraud - and suffering when they make the wrong decisions - this new community is giving the originating and terminating networks the power to tell if calls are fraudulent, and hence the power to stop them on a case-by-case basis.
Instead of crudely blocking a number range or even a whole country, the telco can verify whether a call is legitimate or not by exchanging information with another member of the AB Handshake community whilst the call is still being set up. If a call looks like it originated from a particular network, but the 'handshake' response from that network says that they know nothing about the call, then the terminating operator knows they can safely block the call because somebody is misleading them about the call's origin. Similar handshake checks prevent call stretching and other manipulations. The process of exchanging handshakes and checking them for consistency takes just a few milliseconds, meaning customers will not notice and the risk of paying money to fraudsters is eliminated.
Returning power to the communications providers at either end of the call also means they can choose to connect the call even if there is a mismatch in the handshake information from the originating and terminating operators. Telcos can decide their own priorities when it comes to maintaining continuity of service and inhibiting fraud, instead of relying on wholesale carriers to make these important decisions for them.
As more telcos adopt the AB Handshake to authenticate calls, so they each enjoy more comprehensive protection. Our community is growing. Let's build the fraud-free community together.
