The Kairos Fraud Detector System

The Kairos Fraud Detector System (hereinafter referred to as KFD or the KFD System) is a solution designed to identify and analyze user behavior during sessions in both web applications (accessed via a browser on a personal computer or mobile device) and mobile applications that provide access to electronic banking. KFD incorporates algorithms to assess the probability of fraud based on attributes retrieved from communication protocol headers, network parameters, and the status of the mobile device or web browser.

KFD features an open, highly flexible architecture composed of multiple modules, each with specific functionalities for preventing fraud and unauthorized transactions in real time. One of the main modules focuses on detecting potentially dangerous applications, with particular emphasis on remote desktop applications, as well as detecting malicious plug-ins or packages appearing on blacklists (commonly used to gain unauthorized access to user login data in mobile banking). Screen sharing is especially dangerous for mobile devices because they typically rely on on-screen keyboards, making it easier to capture entered login data.

Data Collection and Processing

When a user attempts to log in to electronic banking, the first phase of the data flow (data collection) is initiated. At this stage, a comprehensive set of information about the end device (the device on which the user session is maintained) is generated and transmitted—indirectly through the backend of the Bank’s Central System—to the KFD System. KFD is compatible with both mobile devices and personal computers.

For web browsers, detailed parameters that may indicate a fraud risk are gathered using a JavaScript library, which is interpreted by all supported browsers. Based on these collected parameters, the KFD System backend generates specific attributes, which are then used by the KFD event-generation engine for further evaluation.

The basic functionality of the KFD System also includes support for mobile applications. The methodology for collecting attributes that describe the end-device status (including operating parameters at the device, system, and application levels) is outlined in more detail later. As with browser-based data collection, a dedicated library is made available for mobile applications, containing the necessary functions and methods to gather a complete set of end-device status attributes.

At this stage, data about the individual end-user device is already sent to the KFD System. This includes operating and configuration parameters of the user’s device, as well as other information used to determine two unique identifiers:

Finger Print – an identifier generated for the specific device instance.
Digital Foot Print – a set of attributes characterizing the device’s network environment, including IP addressing, a list of paired Bluetooth devices, GSM cellular network data, and location data.

The actual data collected depends on the user’s permission settings. Methodologies developed by Momentum Solutions on the KFD side then trigger rating algorithms to determine the fraud risk level. The unique logic used to interpret the end-device status attributes allows KFD to carry out this risk assessment on the fly —meaning that a warning signal can be triggered even before the user finishes logging in or starting a new session within the web browser or mobile application. Consequently, KFD enables fraud detection at the moment of its initiation, preventing a fraudster from successfully implementing unauthorized activities.

Extending Detection Beyond Login

KFD effectively prevents account takeover (ATO) resulting from stolen, intercepted, or even voluntarily shared login data. The KFD System enhances the onboarding process, as well as KYC and AML procedures, by introducing invisible (UX-neutral) protective measures. In addition to combating fraud and unauthorized transactions, KFD helps prevent marketing bonus abuse and the illegal trade in bank account login data.

Crucially, fraud risk analysis does not end after the initial login. The KFD System supports the collection, transmission, and analysis of a comprehensive set of status attributes throughout each subsequent user action. Any request sent to the Bank’s Central System backend or mobile application triggers a new round of data gathering. Along with static analyses of individual data sets, Kairos Fraud Detector also tracks changes in parameters such as the Digital Foot Print, location data, or the Finger Print.

Conclusion

The Kairos Fraud Detector System (KFD) enables real-time detection and prevention of fraudulent activities in electronic banking environments.

Kairos Fraud Detector (KFD)MOMENTUM SOLUTIONS

Kairos Fraud Detector (KFD)

MOMENTUM SOLUTIONS

Kairos Fraud Detector (KFD)

MOMENTUM SOLUTIONS

Kairos Fraud Detector (KFD)
MOMENTUM SOLUTIONS