Smarter fraud decisions, made visible
See exactly how our machine learning models assess fraud risk. Get insights into the signals behind each decision so you can act with more confidence and control.
April 2025
Understand the why behind every fraud block
We’ve made it easier to see how fraud decisions are made. Now, when a payment is blocked or flagged, you can view the specific signals our machine learning models used to assess the risk, drawing from your own data, checkout inputs, and our global network.
Make more informed decisions
This transparency helps you evaluate the model’s logic and adjust your setup with custom rules when needed. Whether you're looking to refine your risk settings or provide clarity to operational teams, these insights are now at your fingertips.
Explore the signals
To access the signals, go to any card based payment that was fraud disputable (e.g. a Mastercard payment). You can do so via the Customer Area > Transactions > Payments > Select a payment and click on the Risk score.
To further drive the classification accuracy of Protect, please ensure that you are sending all required fields, where possible including Checkout information to drive the highest possible accuracy.
Latest updates
- Customer Area
Gain deeper insights in onboarding performance with our newest dashboard
Track onboarding performance and spot bottlenecks with the new KYC verification insights dashboard, helping you streamline the process for your account holders.
- Authenticate
Maximize conversion under Japan’s 3DS regulation
Stay compliant and reduce checkout friction with Risk Based Authentication, now live for Japan’s 3DS mandate.