[Beta] Fraud Prevention / Risk Check

For every payment started with Ivy, Ivy's risk scoring engine calculates the probability of fraud for this payment.

How does Ivy's risk-scoring engine work?

Ivy's risk-scoring engine runs on a machine learning model trained on millions of data points from every financial layer of a transaction, ranging from checkout data to detailed user insights through bank account details and enriched payment data.

Data points include:

product, geography, sex, education, marriage, age, repayment, transactionAmount, transactionHistory, identity, creditScore, regularIncome, balance, TC40Records, deviceFingerprint, ipAddress

How does Ivy's risk-scoring engine help me?

Generally, a payment with Ivy has the status waiting_for_payment when the user executes the payment through their bank account. Ivy's risk scoring engine then calculates the probability of fraud for this payment.

Fraudulent Payments

If a payment is flagged as fraudulent, Ivy automatically cancels the payment and notifies the merchant.

Guaranteed Payments

If a payment has a >99.9% probability of success, Ivy will instantly switch the payment topaid even before the money has been settled and guarantees the payment.

Standard Payments

If a payment is neither flagged as fraudulent nor guaranteed, Ivy will update the payment status to paid and notify the merchant as soon as the money has arrived in Ivy's bank account.

Enable risk check per payment

In order to enable a risk check for a new Payment, simply set the field riskCheck: boolean to true in the Checkout Session Create Request.


Enabling a risk check is a billable event. Please check this with your point of contact at Ivy.