Online retailers depend on machine learning to detect fraud
For fraudsters, the holiday season might be the most wonderful time of the year. ZDNet reports that as UK consumer habits during the Black Friday weekend shift from brick and mortar to online shopping, fraud increases, in turn. The outlet reports the rise of machine learning in anti-fraud efforts.
On average, retailers attributed 7.6 percent of annual revenue across all channels to fraud in 2016, according to a report. A more recent analysis from Sift Science, however, found that these peak retail days do not lay claim to the days with the highest proportion for attempted fraud since legitimate attempts dwarf these attacks.
Kevin Lee, Trust & Safety Architect at Sift Science explains that while the number of fraud orders goes up during the holidays, his team found that the ratio of fraud actually goes down during this time of the season. Similar analyses by other anti-fraud experts such as iovation back their findings.
Machine learning is the latest tool that merchants are using to fight back against fraudsters. “We use supervised machine learning, because it ramps up faster than unsupervised -- does not take as much data to learn,” Lee said to ZDNet about Data Sift’s methods. “We leverage our global model -- over 12,000 websites use us and we use them to protect each other.”
“Our machine learning algorithms analyze billions of combinations of inputs to detect subtle fraud trends across multiple businesses and industries more quickly and accurately than a human.”
Iovation, another company that tackles fraud prevention, also leverages machine learning to spot fraud attempts. “Fraud analysts can catch certain types of fraud machines might miss and react quickly to threats that are unique to that business. But they may miss trends that are too subtle for humans to pick up on or are only noticeable on a global scale,” Angie White, a product marketing manager at iovation, explained to ZDNet. “Our machine learning algorithms analyze billions of combinations of inputs to detect subtle fraud trends across multiple businesses and industries more quickly and accurately than a human.”
The online retail anti-fraud business will soon have to contend with new EU regulation called PSD2 to go in effect in 2019. ZDNet reports that online businesses with high fraud occurrences will now subject transactions from €30 and up to heavy authentication, in an effort to curb fraud. Anti-fraud experts appreciate the intention, but see this measure is excessive.
White points out the extra steps could deter potential customers. “Conversion rates are already low in this space, and any added obstacles or friction could correlate into an increase in cart abandonments.”