Machine Learning continues to Combat Financial Cybercrime

Over the past few months, unemployment systems in America have been continuously scammed by sophisticated international hackers, who exploit loopholes in the government online system in order to gain government unemployment benefits multiple times.

The fraudsters use Personally Identifiable Information (PII) stolen from online sources, such as names, addresses, social security numbers, phone numbers et cetera to truck public servants into accepting their illegitimate unemployment claims.

The government payouts are then sent to accomplices that launder the money in order to mask the illegal nature of their acquired payments. Current rules-based systems used to spot scams using illegally acquired PII are simply inadequate, but AI using high-resolution data is quickly proving more than adequate in recognising cybercrime scams.

Companies already use AI to monitor emails, website logins, personal and business transactions daily in order to log suspicious activity and potentially flag atypical consumer behavior in order to catch cybercrime as it happens, and is now being used to spot more sophisticated scams such as government unemployment benefit false claims.

The AI process, called anomaly detection, has rapidly become the go-to machine learning technology used to catch unusual account activity, and is now being implemented in more and more software architecture to catch cybercrime as it is being committed.

In the constant tug of war between hackers and cyber security, the implementation of AI and machine-learning has proven very effective, allowing anomalous behaviour to be flagged as it happens across many different online targets to the point that emerging patterns of behaviour can be pointed out to cyber security experts before they themselves discover it.