Please note: The algorithm descriptions in English have been automatically translated. Errors may have been introduced in this process. For the original descriptions, go to the Dutch version of the Algorithm Register.

Automated document check and face comparator.

Automated document check and face comparator when registering in the Basic Registry of Persons (BRP) and applying for identity documents for persons over 18 years old. This Algorithm helps the registrant prevent "look-alike fraud". Sourced from the Rijksdienst voor Identiteitsgegevens (RvIG).

Last change on 2nd of February 2024, at 7:56 (CET) | Publication Standard 1.0
Publication category
High-Risk AI-system
Impact assessment
Field not filled in.
Status
In use

General information

Theme

Organisation and business operations

Begin date

2021-07

Contact information

Contact - Gemeente ’s-Hertogenbosch

Responsible use

Goal and impact

The automated document check and face comparator serve three purposes, namely: 1. Establishing the identity of the person applying for an identity document to prevent identity fraud; 2. Establishing the authenticity of the source documents supplied from the country of origin to prevent forgery; 3. Keeping the information in the Basic Registration of Persons (BRP) reliable.

If the applications using the algorithm indicate that the person whose identity document and face have been scanned is not the person standing at the counter, the Public Services counter clerk engages a colleague to verify this finding. A number of applications are handled by foreign specialists. These specialists also apply the 'four-eye principle' when in doubt about the document and/or face scan finding. If they come to the conclusion that the finding of the document check and/or face comparator is correct and the identity of the applicant does not match the identity on the source documents and/or photograph, the application for registration in the BRP or the application for an identity document is rejected. This rejection may indicate whether fraud is suspected. In that case, Police and/or Marechausse are called in for further investigation.

Considerations

The deployment of this algorithm helps employees of the Public Service Department determine the identity of citizens at the counter. Documents contain a number of authenticity features that the document scanner can determine more accurately than the human eye. This also applies to checking biometric features in the face.

Human intervention

Yes. There is no distinction in the application of the document scanner and face comparator. Both instruments are used for every citizen who comes to register as a resident of the municipality of 's-Hertogenbosch and/or applies for identity documents. Furthermore, the 'four-eye principle' is applied if the finding of the document check or face comparator is negative. Finally, when identity documents are issued, a final check is carried out by a desk clerk.

Risk management

By always deploying the applications when registering for the BRP or when applying for identity documents, the risk of possible profiling is excluded in advance. The next risk is that the automated document check or face comparator leads to an automated decision. This risk is overcome because the finding is always tested by an employee. In case of doubt by at least 2 employees.

Legal basis

Legal obligation to establish proper identity: BRP Act Art. 2.67 4th paragraph, BRP Decree Art.35 Admissibility of biometrics: UAVG Art. 29, in the present situation authentication is necessary.

Links to legal bases

Wet BRP art. 2.67 4de lid, Besluit BRP art.35 Toelaatbaarheid van biometrie: UAVG art. 29: https://wetten.overheid.nl/BWBR0033715/2023-10-01#Hoofdstuk2

Operations

Data

Biometric features in the face and authenticity features on the document.

Technical design

Deep learning. Advanced face recognition APIs: C++, Java, Microsoft .NET, BioAPI 2.0 Verification Engine (C API). Documented examples for key use cases and specific implementations. Tools for biometric evaluations, including generation of 'identification match' views and 'similarity matrix' data.

External provider

Oribi B.V.

Similar algorithm descriptions

  • Automated facial comparison for first registration in the Basic Registry of Persons (BRP), resettlement applications and naturalisation applications for persons over 18 years old. This algorithm helps the registrant prevent "look-alike fraud". Sourced from the Rijksdienst voor Identiteitsgegevens (RvIG).

    Last change on 20th of August 2024, at 7:18 (CET) | Publication Standard 1.0
    Publication category
    Impactful algorithms
    Impact assessment
    Field not filled in.
    Status
    In use
  • Algorithm that can help registrants avoid "look-alike fraud". Sourced from the National Authority for Identity Data (RvIG)

    Last change on 16th of August 2024, at 9:46 (CET) | Publication Standard 1.0
    Publication category
    High-Risk AI-system
    Impact assessment
    Field not filled in.
    Status
    In use
  • Algorithm that can help registrants avoid "look-alike fraud".

    Last change on 13th of January 2025, at 8:11 (CET) | Publication Standard 1.0
    Publication category
    Other algorithms
    Impact assessment
    Field not filled in.
    Status
    In use
  • This algorithm supports RNI staff in preventing "look-alike fraud" by comparing the face of the person at the counter with the image in the scanned identity document.

    Last change on 1st of November 2024, at 16:51 (CET) | Publication Standard 1.0
    Publication category
    Impactful algorithms
    Impact assessment
    Field not filled in.
    Status
    In use
  • The algorithm identifies personal data and pre-entered words in documents. An employee must go through the document and check whether the alert is justified and approve or reject it. An employee can add further markings himself. After approval by the employee, all approved alerts and markings are blacklisted.

    Last change on 14th of October 2024, at 10:02 (CET) | Publication Standard 1.0
    Publication category
    Other algorithms
    Impact assessment
    Field not filled in.
    Status
    In use