Detect risks in customs declarations on entry and import of products

Customs

This algorithm helps Customs and its enforcement partners select goods for inspection based on risk. It uses declaration data from companies and considers whether or not there are risks around product safety.

Last change on 5th of May 2025, at 20:56 (CET) | Publication Standard 1.0

• Publication category: Impactful algorithms
• Impact assessment: Field not filled in.
• Status: In use

General information

Theme

Public Order and Safety

Begin date

01-2009

Contact information

[e-mail]

Link to publication website

https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/douane_voor_bedrijven/veiligheid_gezondheid_economie_en_milieu_vgem/vgem_algemeen/vgem_algemeen_2

Responsible use

Goal and impact

The purpose of this algorithm is to select consignments for inspection that enter the European Union, through the territory of the Netherlands, and that are imported into the Netherlands. The algorithm focuses on consignments containing products. The legal basis is Regulation (EU) 2019/1020 on market surveillance and conformity of products. Here, Customs acts in cooperation with the Netherlands Food and Consumer Product Safety Authority (NVWA) and the Netherlands Labour Inspectorate (NLA). Customs checks companies more or less based on the results.

Considerations

Customs processes 700 million declarations every year, including in the field of product safety. With this algorithm (risk profiles), Customs can assess declarations more efficiently and faster than without using this algorithm. Thus, declarations are approved or checked faster. And goods are released faster. Also, the algorithm ensures that 'known and reliable' companies are actually checked less.

Human intervention

If a declaration produces a match with a risk profile, a Customs employee reviews the declaration based on the algorithm. If necessary, the Customs employee asks for more information. The decision to approve a declaration, or instead to start checking, is made by a human being.

Risk management

A risk profile is always checked by a second customs officer (4-eye principle) before use. And once the profile is in use, it is periodically checked how many declarations produce a match. If necessary, the profile can be adjusted. Annually, Customs evaluates per profile whether it is necessary to refine, extend or terminate it. Customs monitors internal and external complaints and incidents, for example if a profile generates unjustified raking and therefore designates too many declarations as higher risk.

Legal basis

The legal basis for food safety controls is in the Union Customs Code (DWU), the General Customs Act, Regulation (EU) 2019/1020 and the Commodities Act. The use of risk profiles is also stipulated in Article 46 of the DWU. This says customs controls must rely primarily on risk analysis by electronic data processing techniques.

Links to legal bases

• Douanewetboek van de Unie (DWU): https://eur-lex.europa.eu/legal-content/NL/TXT/?uri=CELEX:32013R0952
• Algemene Douanewet: https://wetten.overheid.nl/BWBR0023746/2016-07-01
• Verordening (EU) 2019/1020: https://eur-lex.europa.eu/legal-content/NL/TXT/?uri=CELEX%3A32019R1020
• Warenwet: https://wetten.overheid.nl/BWBR0001969/2023-04-19

Operations

Data

The algorithm can use any data from the goods declaration. For example, type of goods, goods code, weight, number of packages, consignee, declarant, consignor, country of consignment, EORI number, means of transport, etc.

Links to data sources

Gegevensvereisten douaneaangiften (zie Bijlage B Gedelegeerde Verordening Douanewetboek van de Unie + Bijlage VIa van de Algemene Douaneregeling): https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/douane_voor_bedrijven/naslagwerken_en_overige_informatie/andere_onderwerpen/gegevensvereisten/

Technical design

The algorithm consists of decision rules created in collaboration with content experts. These decision rules provide estimates of which declarations have a higher risk of entry and entry. These decision rules are based on 'if-then-else' combinations.

An example: if (if) the information we have on a shipment shows that the goods have an abnormal weight, (then) the declaration is checked manually and the shipment may be selected for extended inspection. If the goods do not have an abnormal weight (else), we do not stop the shipment. Or at least not based on this criterion.

External provider

Internally developed