**Last updated**: 30 September 2025 | [**Change log**](/products/fraudsight/changelog/)

# Get started

Minimize fraud and increase approval rates using the latest in data insights. Use an [integrated risk assessment](/products/payments/) as part of the Payments API or follow the steps below for a separate risk assessment request.

**Supported payment methods**: *American Express / Carte Bancaires / Discover / Diners / EFTPOS / ELO / JCB / Maestro / MasterCard / MasterCard Debit / Visa / Visa Debit / Visa Electron (UK only)*

Prerequisite
Make yourself familiar with our [API Principles](/products/reference/api-principles) to ensure a resilient integration.

## 1. Set your headers

Use our [API Reference](/products/fraudsight/openapi) to set your headers.

## 2. Collect device data (optional)

This step gathers customer browser details and performs a GeoIP lookup. It adds more data points for the risk assessment or manual rule creation.

Web device data
ThreatMetrix JS library for web sites

Android SDK
ThreatMetrix Android SDK (coming soon)

iOS SDK
ThreatMetrix iOS SDK (coming soon)

## 3. Risk assessment (required)

This is the API request containing payment, customer and order information. Optionally, you can provide a `sessionId` from the Device Data Collection (DDC).

Risk assessment
Risk assessment using details from the payment method, customer and order

SCA Exemptions
You can request an SCA exemption assessment as part of the FraudSight request. This is also available as a separate [exemption request](/products/sca-exemptions/).

## 4. Payment outcome and fraud details (conditional)

When using our [Card Payments API](/products/card-payments) for the payment itself, the outcome of the payment (e.g. authorized) and any subsequent fraud or chargeback information is automatically fed to the fraud data model to improve assessment accuracy. For any other API processing the payment (Worldpay or 3rd Party), the following steps are required:

If you're using other gateways/acquirers for the payment gateway you can still use FraudSight but you must provide additional details so the data model can learn and mature for future risk assessments.

Payment
Provide details as to whether the payment was authorized or not.

One or both of these should also be provided:

Fraud
Using TC40/Safe data, did fraud occur on any risk assessed payments.

Chargeback
Did any of the payments result in any fraud specific chargebacks.

**Next steps**

[Device Data Collection](/products/fraudsight/device-data)