Page 8 - GS180601

ReadersSpeak

Machine learning can safeguard global payments When it comes to CNP transactions, honest players around
the globe – genuine consumers, merchants, processors and
ince the introduction of EMV, fraudsters have financial institutions – want two things: speed and secu-
focused on card-not-present (CNP) transac- rity. If an imbalance exists on either end of the spectrum,
tions. A February 2018 study by Javelin revealed the entire transaction is jeopardized, resulting in a lost
S fraudulent transactions are 81 percent more sale to the merchant and frustrated consumer. Existing
likely to occur online than at the POS. This is likely fraud models capture patterns and create alerts that sig-
attributable to the rise in online merchants and the rise of nal transactions that could be fraudulent. The challenge
digital transactions. As, globally, we migrate to shopping behind these indicators is that they sometimes miss new
online more frequently, fraudsters continue to find ways fraud because they are not self-taught. Machine learning
to exploit loopholes in transactions. This is where machine solves for this challenge, by understanding the customer
learning moves from a buzz word into practice. at each individual touchpoint and constantly addressing
new fraud patterns. This helps eliminate false positives,
Machine learning models can use the dollar amount of the ensuring the right transactions do go through, and rec-
transaction, the location and device from which a transac- ognizes new fraud attacks as they occur. This delivers a
tion is made, along with hundreds of other data points – risk- and friction-free experience and is the only way both
all in real time – and establish patterns in a consumer's be- parties can be satisfied. The increasing reliance on speed
havior. Each time a purchase is made, the machine learn- and security leaves global payments vulnerable to fraud.
ing model compares the most recent transaction char- Machine learning provides that extra layer of protection.
acteristics to the historical profile to determine, within Dave Excell
milliseconds, if the transaction is legitimate or suspicious. Co-founder and CTO of Featurespace Ltd.
This ensures that the highest level of security is provided,
mitigating risk, eliminating friction and allowing more Have insights to share?
genuine transactions to be processed. Many thanks to Dave Excell for sharing his expertise
here. We will welcome your helpful perspectives at
greensheet@greensheet.com.

What you see isn’t

always what you get.

Transparency builds better

relationships, and you can’t
expect to grow your business

with anything less. We believe in
ownership through partnership.

Created by ISOs. Owned by ISOs. Serving ISOs.

WWW.COCARD.NET 800.882.1352