Back to all updates

Identify up to 2x more fraud without increasing casework

Score Premium platforms can now use a new fraud model that identifies up to 2x more fraud while reducing false positives, helping your risk team improve fraud detection without creating additional casework.

September 2026

A close-up of an Adyen payment terminal with the company's logo on the device.

Detect more fraudulent behavior

The new acquiring fraud model uses more than 900 features and updated fraud labels to identify fraudulent behavior more effectively. It replaces the existing "Account flagged as very high risk" model and is available in Score under "Account behaviour matches known fraud."

Improve fraud detection with fewer false positives

The new model increases detection performance while reducing false positives, so you can identify more fraud without creating additional work for your risk team. You can configure the signal using the same tools as other Score signals.

Review the new signal in Score

Review the new "Account behaviour matches known fraud" signal in your risk rules and assign an automated action if you want to use it. This update is available to platforms using Score Premium.

Learn more

Identify account holder risk

Go to Adyen Docs

Latest updates