How many indicators of fraud do your fraud analysts track? Over how many billions of computations are you monitoring for fraud indications? A lot, right? Can Artificial Intelligence (AI) help you track systemic fraud faster? This week’s Outlier of the Week shows how AI tracks fraud.
Credit card issuers need to track transactional and systemic fraud daily. It’s billions of data points for fraud analysts and multiple software programs to process and analyze. Using Outlier, a credit card issuer found a chain of gas service stations had a lot of inconsistency in their data. While analyzing the gas station credit card transactions, Outlier’s Automated Business Analysis platform found an unusual spike in activity and alerted the credit card fraud team of this unexpected change.
Reading the Story That Used AI to Track Fraud
Seeing from the story above, the gas station category saw a spike to 4.7K transactions, which is a month-over-month increase of 5% and is 5% higher than the expected model. The population impacted did not change compared to the previous month’s population, remaining at 0.1% of Entries. Outlier’s Root Cause Analysis feature shows that the spike was due to new credit cards being used in the state of Texas from one specific merchant.
With this Outlier story, the fraud analyst team had a quick lead on a possible systemic fraud, which was being masked by running new cards out of a few gas stations in Texas. This story saved the fraud team hours to weeks of investigation into this crime ring.
You can identify meaningful trends in your data with Outlier, an automated business analysis platform. Outlier empowers businesses to take a deeper look at their data and uncover unexpected trends that can lead to revenue increases. To see how Outlier’s AI tracks fraud, sign up for a custom demo.