Employing AI, ML in Customs and border management

	Vietnam, Cambodia to enhance ties in border management, defence
	China tightens border management, Hai Phong recommends exporting enterprises
	Customs Engaging Traditional and New Partners with Purpose

Speakers attends a panel session at the 2023 Technology Conference and Exhibition of the World Customs Organization (WCO) co-hosted by The General Department of Vietnam Customs in Hanoi on 10th – 12th October 2023

Benefits of AI and ML application to Customs operations and border management

According to the WCO’s report, use of big data, data analytics, AI and ML in Customs and border management presents a tremendous opportunity in the cross-border movement of people and on the commercial side.

As huge volumes of data are generated by people and goods moving across borders, this group of technologies provides the ability to make sense of this vast and ever-increasing amount of data.

These technologies can be used to ingest this data and detect and predict patterns more accurately than humans can. Visual search and facial recognition technology, and behavioral and predictive analytics, which are already being used in other sectors, can also be further tailored for use in Customs and border management.

AI can be used for the following purposes: developing revenue collection models, ensuring that the appropriate duties and taxes are collected at the border; classifying products under the Harmonized System (HS), simplifying matters for users and enabling greater compliance and certainty for both Customs and the private sector; supporting the Customs audits to identify anomalies much more quickly and thereby enable Customs auditors to focus on areas of non-compliance; improving risk-based targeting of commercial shipments, as well as to provide and analyses data during shipment inspections using augmented/mixed-reality glasses in detecting contraband and counterfeit goods; analyzing container images made by X-ray scanners to improve the efficiency of cargo inspection; monitoring and control logistics in Customs warehouses and bonded areas; identifying high-risk passengers and vehicles by using visual search and facial recognition technology at the border.

This can be further expanded to create intelligent analytics to predict future outcomes, facilitating better risk management and preparedness; providing better service by placing enquiry robots for passengers at the border; and providing Customs duty self-payment services by developing mobile apps at the border, etc.

The WCO has developed a number of tools to support Members in developing their data analysis and data analytics skills.

For example, the “Handbook on Data Analysis” (2018) is aimed at presenting a highlevel overview of data analytics and, more specifically, what it is, how it works, and how useful it may be to Customs and other governmental agencies.

The Handbook offers guidance on how to leverage big data and data analytics, outlines data governance-related issues, and presents some common data analysis tools, such as predictive analytics, cognitive computing and statistical programming languages.

In addition to the “Handbook on Data Analysis”, the “WCO Capacity Building Framework for Data Analytics” is designed to help Customs officers develop their organizational and technical capabilities to capitalize on the rapid growth of collected data and gain insights to make better-informed, data-driven decisions.

Using AI and ML technologies of Customs administrations around the world

In the WCO regional workshops on disruptive technologies held during the course of 2021, 2022 and 2023, some Members shared their experience on the use AI/ML in Customs.

For example, Botswana Customs implement online payment through a mobile app (BURS Mtax) which provides a number of benefits to the trading Community as well as the revenue authority. The Zambia Customs “TaxOnApp” is another example of this type of implementation.

The Nigeria Integrated Customs Information System (NICIS) Smart Fraud Detection is an AI/ML technology designed to assist Nigeria Customs Service Officers in performing risk analysis on the declaration’s information, processing a huge amount of data in the shortest possible time.

Azerbaijan Customs are working on a project on declaration processing using AI/ML. El Salvador Customs has created an “Artificial Intelligence Department” that works on improving a number of processes by implementing AI, such as risk management, image processing and valuation of goods.

Indonesia Customs uses AI/ML for passenger risk management and import risk management purposes.

Russian Federation Customs also uses AI for risk management purposes. Japan Customs utilizes AI for risk management and post-clearance audit purposes.

At the 2023 WCO Technology Conference and Exhibition taking place in Hanoi, Vietnam, Mr. Philippe DUPONTEIL, Director of European Commission, Directorate General Customs and Taxation, Digital delivery of customs and taxation policies; Customs tariffs shared AI employment in Customs operations.

AI has many different practical applications in customs operations and can bring many benefits. Accordingly, AI is applied in risk management and predictive analysis. AI can certainly analyze larger volumes of data than humans and can identify unclear trends.

AI can not only analyze large volumes of text data (such as customs declarations), but also analyze images and compare them with the content in customs declarations.

Besides the application of AI in risk management and predictive analysis, AI can also be used to promote the development of application software in the customs sector and for training purposes. For example, at the European Commission, AI is used to create e-learning models on various topics related to customs and tax operations. Usually, a script will be prepared, then AI will create a digital teacher with a voice to present and teach.

World Customs Organization welcomes new Secretary General

However, we should not depend entirely on AI at work. Because AI can make inaccurate decisions. These decisions can affect management agencies, businesses and people, and lose their trust, thereby causing loss of budget revenue.