How AI Is Winning the War Against Cyber-Criminals

Thursday, December 21st, 2017

One of AI’s greatest innovations today focuses on solving the businessman’s biggest problem: fraud. With the rapidly increasing rate of transactions, detecting fraud means keeping up with today’s tricksters who are constantly out to discover and exploit new points of weaknesses. And in most cases, the adversary always has the odds of winning— they are often more motivated, well synchronized, and smart.

In his recent report on Fraud, Pedro Bizarro, the Chief Science executive of Feedzai, an AI Company reports how difficult it has been to keep up with the rate of fraud. Cybercriminals are making use of up-to-the-minute technology, for instance, launching bot attacks that are capable of adding to a cart five times faster than humans, and often target valuable merchandize or newly launched items.

Fraudsters take cover on the huge piles of records and innocent-looking hiding places. Data science experts at Feedzai have realized tricksters use null areas in device IDs, and this has contributed to 78% fraud rate. This is because cybercriminals use emulators to disguise names linked with their devices.

Customers on the other hand continue to demand immediacy, and we may see payments go more and more real-time. But how can a business make excellent decisions amid the risky, fragmented, and speedy landscape? Well, Artificial Intelligence brings merchants a paradoxical promise: perhaps the most ultramodern tech that can shift state of the economy to the old days when online business was a one-on-one consumer experience.

Ecommerce grew rapidly because business of the surety about transactions because of real-time nature of the trade. These days, the word “trust” has no meaning in the online business world as all sorts of fraud have infused the industry, some techniques even untraceable.

So the key to a successful fight against fraud requires compressing the large piles of raw data into comprehensive and accurate insights on all transactions as they take place. AI’s idea of making big insights out of big data is a product of theoretical physics. Physicists understand complex things by breaking them down into their primary components. However, there’s also the technique of going big by combining primary components to the see patterns that emerge.

Andrew Tikovsky, Data Science expert at Feedzai, describes it as “combining simple things to get new patterns”. Feedzai develops their chargeback protection tech out of simple features/components. Because it is possible to tell the time when someone initiates a transaction and whether they have really been in a physical place or not, we can combine such data together and come up with a concept of how they interrelate in a fiscal ecosystem.