AI Case Study

Paypal keeps fraud losses at about one-third of industry average using machine learning and deep neural networks to predict whether a transaction is fraudulent or normal

Paypal's artificial intelligent platform uses deep learning and machine learning to update rules to map customer spending patterns and determine whether a transaction fits the pattern.

Industry

Technology

Internet Services Consumer

Project Overview

"When a spending pattern is revealed—for example, if sudden strings of many small purchases at convenience stores turn out to be fraud—it’s turned into a “feature,” or a rule that can be applied in real time to stop purchases that fit this profile. Thousands of ‘features’ are processed now, compared to hundreds when the system was first put to use in 2013.

As a result, PayPal can now do things like tell the difference between friends buying concert tickets together and a thief making similar purchases with a list of stolen accounts. And it’s all done in-house to avoid even the tiny latency that would occur if the company relied on a cloud provider. “Thousands of ‘features’ searching through 16 years of users’ history all needs to be done in less than a second,” Wang says."

Reported Results

Fraud rate of .32% compared to industry average of 1.32%

Technology

Function

Risk

Security

Background

"From a cybersecurity perspective, ­PayPal has a target on its back: it processed $235 billion in payments last year from four billion transactions by its more than 170 million customers. Fraud is always possible via theft of consumer data in breaches such as “phishing” e-mails that con users into entering their credentials. To keep ahead, PayPal relies on intensive, real-time analysis of transactions."

Benefits

Data