Financial Crime World

Credit Card Fraud Prevention in Switzerland: How AI and Human Intelligence Work Together

Switzerland’s financial sector is home to a sophisticated software system that has been successfully preventing credit card fraud at UBS Card Center for over two decades. The FICO Falcon platform, which assesses 500,000 transactions daily - equivalent to 25% of all credit card transactions in the country - uses artificial intelligence (AI) and machine learning to identify suspicious activity.

How it Works

The platform has been fed with a myriad of rules and behavior patterns, allowing it to analyze the volume, type, and amounts of all transactions. AI also examines card use at branches and other locations, triggering a system warning if it detects an inconsistency.

Machine Learning in Action

  • Compares transactions against model profiles
  • Identifies suspicious activity based on various rules and known practices used by scammers
  • Analyzes transaction volume, type, and amounts

However, Anna from UBS Card Center notes that not every suspicious case proves to be true. For example, a customer was flagged for making multiple small purchases within a short timeframe using NFC and no PIN - only to reveal that she was simply stocking up before a vacation.

The Importance of Human Intelligence

Human intelligence plays a crucial role in the fraud prevention process. The UBS Card Center’s fraud team works daily to improve the accuracy of the early warning system, creating new rules and learning how to interpret behavior patterns correctly based on employee input.

Human Expertise

  • Processes potential cases of fraud
  • Creates new rules and improves interpretation of behavior patterns
  • Achieves an 84% increase in fraud prevention rate between 2015 and 2018

As Marcel Drescher, Head Fraud Services at UBS Card Center, notes, “Our main expertise is interpreting and analyzing individual cases. The system only learns effectively if we feed it the right data.”

The Impact of AI and Machine Learning

The use of AI and machine learning has been instrumental in combating credit card fraud, which saw a continuous rise from 2000 to 2016. However, since 2017, the trend has stabilized due to various security mechanisms and technological developments increasing security for credit card users.

A New Era of Security

  • Credit card fraud prevention rate increases by 84% between 2015 and 2018
  • Continuous monitoring and improvement of the early warning system
  • Enhanced security measures and technological developments stabilize fraud rates since 2017