Fraud Intelligence Integration
Fraud Intelligence is a Saferpay module that protects merchants from fraudulent online transactions. It relies on industry-leading AI technology and allows merchants to dynamically react on suspicious behavior and even prevent transactions with malicious intent, during authentication.
This chapter will cover the technical aspects on how to integrate Fraud Intelligence in your application.
Requirements
A corresponding license and thus a valid identification with a username and password for the Saferpay system.
Availability of at least one active Saferpay terminal via which payments can be carried out, and availability of the associated Saferpay TerminalId.
A contract to use the Fraud Intelligence module in Saferpay. Please contact your contractual sales contact on that matter.
Saferpay JSON API SpecVersion 1.20+
Supported Payment Methods and Flows
Currently, the following payment methods are supported:
Visa/VPay
Mastercard/Maestro
American Express
Bancontact
Diners Club
JCB
PayPal
Unionpay
Currently, the following flows are supported:
Transaction Interface, including Saferpay Fields
Data points
In order for the fraud detection to work properly, the system needs to be provided with a set of data points with each transaction. Some are provided automatically by Saferpay, while others need to be submitted by the merchant's system with the initial request, when starting the transaction with either Transaction Initialize or Payment Page Initialize.
All of these datapoints are generally optional. However the detection will work better the more data are provided.
The following data points can be set via the JSON API:
Deprecated Fields
This fields are not in use anymore:
Device Fingerprinting
If Devide Fingerprinting is enabled, the following additional data points are available:
Example
Here you can see an example Payment Page Initialize request. Note that the containers and parameters are, of course, consistent throughout the whole API:
Device Fingerprinting
Device Fingerprinting is a process to identify and track devices across websites through a central service. The data collected is used to identify malicious devices and thusly to prevent fraud.
The Saferpay Fraud Intelligence solution does offer possibilities to benefit from such a process, however it needs additional integration steps, in order to function.
Preparation
Before you can start integrating, please contact dl-fraud-team-saferpay@worldline.com in order to get your so called instanceId, which is important for this service to function.
Integration
The integration of the Device Fingerprinting service is done on client-side via JavaScript.
It is imperative, that this script is called before the Saferpay payment is initialized.
Code Sample
Explanation
The above script creates the fingerprint itself and directly submits the data towards the Fraud Intelligence service.
dfpBaseUrl: This URL static and provided by us. It will change shortly and we'll then inform what the new URL is.
session_id: A unique id generated by you which is diferent for each transaction. It is also used by Saferpay, see the next step.
instanceId: The instanceId which was given to you by the DFP support
pageId: This value always needs to be "P" which stands for "Prevention" and helps to route the data to the right location.
Saferpay Integration
Once the above script has been successfully initialized and the necessary device-fingerprint has been created, you can initialize the Saferpay Payment.
It is however important, that you share the unique session_id
you previously created with Saferpay. You do so, by submitting it inside the RiskFactors.DeviceFingerprintTransactionId
parameter, via either the PaymentPage Initialize, or Transaction Initialize request, depending on your integration.
Risk Score
The Fraud Intelligence module uses artificial intelligence to calculate a Risk Score from all available data. This Risk Score is an indicator how risky the transaction is: A transaction with a high Risk Score generally has a higher propability to be fraudulent than a transaction with a low risk score.
Rules
Rules are used to configure the individual risk appetite of a merchant and to make a decision based on the calculated Risk Score and other criteria. For example, a simple rule could block all transactions with a Risk Score higher than x. A rule can also combine multiple conditions such as velocity thresholds and check other data points, which allows to model complex business requirements and manually block specific fraud scenarios.
Responses
Success
In case of a success, the transaction response will also carry additional information inside the FraudPrevention.Result parameter. This can have one of two values: APPROVED and MANUAL_REVIEW.
In both cases, the transaction was indeed successful. However, the latter indicates that there may be issues with this transaction, which need to be reviewed manually.
It is then up to you, the merchant, to either accept or decline this transaction.
Failure
In case of a decline, Saferpay will throw a appropriate error, also containing the reason.
Feedback to Fraud Intelligence Hub
These responses will in return be transferred to the Fraud Intelligence Hub for further Risk analysis and overview purposes.
Note, that this currently only happens for normal debit transactions. Other transactions, like refunds, are currently not included.
Last updated