/
5 minute read
June 24, 2022

Using the Fraud Feedback API (Beta)

Hidden
Note
This feature is currently in beta and subject to change. It also requires additional activation steps. To learn more about the Beta program for this feature and about activating it for your program, contact your Marqeta representative.

Sending a request to the /feedback endpoint informs the Real-Time Decisioning risk model whether an approved transaction was confirmed to be fraud. For details on the /feedback endpoint, see Fraud Feedback API Reference.

Handling false positives

Use the AMDL for a state and a rule to leverage the Fraud Feedback API call for handling false positives:

JSON
Copied

Is this helpful?

Yes
No

The Fraud object

The following is the fraud object that contains fraud information.

JSON
Copied

Is this helpful?

Yes
No

The following shows the sequence and an example of the request to inform the Real-Time Decisioning risk model:

  • An authorization transaction in initiated.

  • A transaction comes into Marqeta.

  • The transaction is declined because of a Real-Time Decisioning rule is triggered.

  • Using the Fraudfeedback API, send a request. For example:

    JSON
    Copied

    Is this helpful?

    Yes
    No
  • The next transaction coming in from the same user and card will then not be denied through Real-Time Decisioning.

Join our developer newsletter