You are here
Cybercrime: finding snakes amid the butterflies
THE way to understand how suspicious transactions can be missed by banks is to think about a system that has been trained to identify millions of butterflies.
But while that system, akin to a digital lepidopterarium, is perfectly able to identify positive cases, it will not be able to pick out scorpions or snakes because it simply hasn't been trained to do so.
So banks' instructions to the system is to "show me everything that is not a butterfly", says Mark Gazit, CEO of Israeli cybersecurity firm ThetaRay.
This results in a tremendous load of anomalies - scorpions and snakes that don't fit normal transaction patterns - that compliance officers at banks have to sift through to pick out the false positives from the mix.
"In the last five years, the world of financial crime has totally changed. Criminals no longer go into branches to shoot and steal money," says Mr Gazit. "And most of the cybercriminals today are not there to deface websites. It is a very lucrative business."
ThetaRay steps in by effectively trying to codify intuition through machine learning, says Mr Gazit. It taps the part of the mind that draws relationships with parameters.
When someone with a salary of S$2,500 a month gets a sudden million-dollar drop into his or her account, or if he cuts cheques or make bank transfers regularly in small, but consistent amounts to a seemingly unrelated party, these are red flags that a system can be trained to pick up.
This is how the brain works, in joining dots to make connections. Or as Mr Gazit says, when looking at a fresh face, the mind does not pick on the exact distance between the eyes. "You just look at the proportion."
With Asia swelling in wealth, ThetaRay is taking calls from banks from places such as China, Hong Kong, Thailand, Vietnam, the Philippines, and Indonesia, that want to tap into the tech firm's artificial intelligence (AI) capabilities.
ThetaRay is already working with Singapore's OCBC on a pilot, which an OCBC spokesman told BT remains ongoing. Its pilot, first announced in 2017, showed that ThetaRay's technology was able to reduce the number of alerts that did not require further review by 35 per cent. The accuracy rate of identifying suspicious transactions increased by more than four times, the bank said. The test was based on one year's worth of OCBC corporate banking transaction data, with the data anonymised.
Mr Gazit also sees Singapore regulators as forward-thinking when it comes to using AI to combat cybercrime. As it is, Mr Gazit was recently appointed to the Monetary Authority of Singapore's International Technology Advisory Panel. The panel advises the Singapore regulator on international developments in fintech. Mr Gazit will be on the panel for the customary two-year term.
With the recent appointment, Mr Gazit, who also chairs an industry association for fintechs in Israel, hopes to also bring more attention to fintech companies out of his country.
He points out that Israel can offer deep-tech cybersecurity solutions to the rest of the world because it is a product of its environment. "Israel has eight million people. It's a beautiful place, but we have to protect ourselves."