Recurring Fraudsters

Product overview

Recurring Fraudsters helps identify clients who may have previously been confirmed to have committed fraud. It uses 1:N facial recognition to compare a client's facial biometric data against previously flagged face embeddings and identify potential matches for review.

Potential matches are reviewed by a human agent before a screening result is recorded.

Recurring Fraudsters should be used alongside other fraud controls, not as a standalone detection mechanism.

Recurring Fraudsters screening


How it works

Recurring Fraudsters screens new verification attempts against identities that have previously been flagged for fraudulent activity within the Business Partner's database, as follows:

  1. A selfie image is captured during identity verification.
  2. The system generates a face embedding.
  3. The embedding is compared against previously flagged identities using 1:N facial recognition.
  4. If one or more potential matches exceed the configured similarity threshold, they are submitted for manual review.
  5. A human agent reviews each potential match using identity attributes such as:
    • Name
    • Date of birth
    • Document number
  6. The reviewer confirms whether the match corresponds to a previously flagged identity or is a false positive.
  7. The screening result is recorded in the CDD report.

Screening results

Each screening produces a single outcome together with a timestamp and, where applicable, references to related verification IDs.

OutcomeDescription
No matchRecurring Fraudster screening completed at [date] [time] UTC and no hits were found.
Match confirmedRecurring Fraudster screening completed at [date] [time] UTC and confirmed Recurring Fraudster hits.
Includes the confirmed matched IDs.

Verification checks

Fourthline performs the following screening during verification.

CheckDescriptionRequired
Recurring Fraudsters screeningCompares the uploaded selfie against previously flagged identities using configured similarity thresholds. Potential matches are reviewed by a human agent before the final result is recorded.

Recurring Fraudsters is supported across all integrations:


Data and privacy considerations

  • Fourthline stores fraud-related personal data in accordance with applicable regulations, including GDPR and local AML requirements.
  • Fourthline stores fraud-related data separately from standard client data and restricts access to authorized personnel.

Support

To enable Recurring Fraudsters, contact your customer success manager.


Get in touch