Antifraud
Getnet Antifraud evaluates the risk of e-commerce transactions by combining customer-provided data with advanced machine learning technologies, neural networks, configurable rules, and manual review by risk analysts.
The accuracy of fraud detection increases as the volume and quality of data used to train the models grow. Richer datasets enable the use of a broader set of fraud indicators, resulting in more precise risk assessments for each transaction.
The Getnet Payment Link API includes an integrated antifraud solution that operates in real time throughout the payment flow. The system analyzes multiple signals—such as purchasing behavior, device information, and data consistency—to generate a risk score and automatically approve or block transactions. When you send complete and accurate customer data as you create the payment intent, the analysis is more effective and overall fraud detection rates improve.
| Field | Type | Description | Example |
|---|---|---|---|
customer_id | String | Unique identifier you assign to the customer | customer_21081826 |
first_name | String | Customer's first name | John |
last_name | String | Customer's last name | Doe Smith |
name | String | Customer's full name | John Doe Smith |
document_type | String | Document type | CPF |
document_number | String | Customer's document number | 12345678912 |
email | String | Customer's email address | customer@email.com.br |
billing_address | Object | Customer's complete address | -- |
The following example shows a customer object in a payment intent:
{
...
"customer": {
"customer_id": "customer_21081826",
"first_name": "John",
"last_name": "Doe Smith",
"name": "John Doe Smith",
"email": "customer@email.com.br",
"document_type": "CPF",
"document_number": "12345678912",
"phone_number": "5551999887766",
"gender": "Male",
"checked_email": false,
"billing_address": {
"street": "Av. Brasil",
"number": "1000",
"complement": "Sala 1",
"district": "São Geraldo",
"city": "Porto Alegre",
"state": "RS",
"country": "BR",
"postal_code": "90230060",
"reference": "Near the hospital"
}
},
...
}{
...
"customer": {
"customer_id": "customer_21081826",
"first_name": "John",
"last_name": "Doe Smith",
"name": "John Doe Smith",
"email": "customer@email.com.br",
"document_type": "CPF",
"document_number": "12345678912",
"phone_number": "5551999887766",
"gender": "Male",
"checked_email": false,
"billing_address": {
"street": "Av. Brasil",
"number": "1000",
"complement": "Sala 1",
"district": "São Geraldo",
"city": "Porto Alegre",
"state": "RS",
"country": "BR",
"postal_code": "90230060",
"reference": "Near the hospital"
}
},
...
}Device Fingerprint
Device fingerprinting is a fraud prevention technique used to uniquely identify the device involved in a transaction based on a combination of technical and behavioral attributes. Instead of relying on a single identifier, the method analyzes multiple signals, such as device type, operating system, browser characteristics, IP address, geolocation, screen resolution, language settings, and other environmental factors, to generate a unique device profile.
Within the Getnet antifraud ecosystem, device fingerprinting helps detect suspicious behavior by identifying patterns such as repeated transactions from the same device using different cards, abnormal purchase frequency, or inconsistencies between device data and customer information. This approach strengthens risk assessment by enabling more accurate detection of fraud attempts while minimizing false positives.
The effectiveness of device fingerprinting increases when device data is collected consistently and correctly during the payment flow. Consistent data produces a more reliable risk score and better real-time decisions to approve or block transactions.
On this page