The rapid expansion of artificial intelligence (AI) across various sectors of the economy and trade is placing countries on a complex trajectory. To obtain optimal benefits from trade, widespread use of AI seems inevitable. Countries are increasingly leveraging AI for trade facilitation, a set of measures to ease the movement of goods across borders, reduce trade costs, and enhance competitiveness. Recognizing the importance of trade facilitation, World Trade Organization (WTO) member countries adopted the Trade Facilitation Agreement (TFA) at the Ninth Ministerial Conference (MC9) in 2013. The agreement entered into force in 2017 and requires member countries to implement trade facilitation measures and fulfill their commitments within a specified period. The core of this agreement is the simplification, modernization and harmonization of import and export processes. This includes a shift from paper-based manual procedures to digital trade processes, with the rise of electronic data exchange, national single window systems, and interoperable digital trade platforms. Advances in AI will accelerate and simplify digital trade facilitation, but it is not without risks.
AI is broadly defined as machine-based systems that can learn from data, generate output, and influence real or virtual environments. A key question is whether and to what extent countries like Bangladesh are prepared to use AI for trade facilitation. The Asian Development Bank (ADB) and the United Nations (UN) Economic and Social Commission for Asia and the Pacific (ESCAP) seek to address this in the Asia-Pacific Trade Facilitation Report 2026: Leveraging Artificial Intelligence in Trade Facilitation. The report, released in the second week of this month, shows that while AI is gradually reshaping trade processes across the Asia-Pacific region, most economies have yet to adopt the technology at scale. The adoption rate of AI in trade facilitation is less than 15% across the region.
Before considering the current state of AI in trade facilitation, it is necessary to understand the overall implementation status of measures in the region. Since its entry into force in 2017, the implementation of the WTO TFA has accelerated efforts towards simplification and harmonization. The United Nations World Survey on Trade Facilitation (UNTF Survey) includes these “TFA Plus” paperless trade measures. According to the study, the implementation rate in the Asia-Pacific region reached about 70% last year, matching the global implementation rate. The implementation rate of digital trade facilitation rose from 33% in 2015 to 60% last year. However, the implementation rate in developed countries is as high as 78%.
The UNTF study covered 62 trade facilitation measures currently in place by countries. The measures are divided into 12 subgroups and cover both binding and non-binding WTO TFA measures, as well as measures that go beyond the scope of WTO TFA, with a focus on digital and sustainable trade facilitation. The five main subgroups of “TFA Plus” measures are: (i) transparency, (ii) procedures, (iii) institutional arrangements and cooperation, (iv) paperless trade, and (v) cross-border paperless trade. Measures based on these subgroups are at the core of trade facilitation.
The remaining seven subgroups are (i) transport facilitation, (ii) trade facilitation for small and medium-sized enterprises, (iii) agricultural trade facilitation, (iv) women in trade facilitation, (v) trade finance facilitation, (vi) trade facilitation for e-commerce, and (vii) green trade facilitation. Most of the measures belonging to these subgroups are highly dependent on the core measures.
According to UNTF research, the average implementation rate of trade facilitation measures in Bangladesh was 70% last year, up from 32% in 2015. The country has made significant progress in transparency, reaching 93%. The progress of the procedure was also satisfactory at 83%. However, Bangladesh lags behind in terms of institutional arrangements and cooperation at 67%. While there has been significant progress in paperless trade (71%), there is still less measurable progress in paperless cross-border trade (33%).
The ADB-ESCAP report identifies five stages of trade facilitation. (i) Paper-based era (before 1990s). (ii) early digitization and electronic data interchange (late 1980s to 1990s); (iii) the expansion of single windows and paperless transactions (late 1990s to 2010s); (iv) emerging technologies (late 2010s to early 2020s); (v) Advanced AI integration (from early 2020s onwards).
ADB-ESCAP conducted a study to gain a clear understanding of the current state, readiness, and challenges of AI in trade facilitation in the Asia-Pacific region. This study focuses on the use of AI by customs and other government agencies involved in trade procedures and complements the UNTF study. A separate working paper presents findings and analysis and provides key input to the joint ADB-ESCAP report.
The research report showed that high-income countries are often better prepared for AI adoption and trade facilitation than low-income countries. Nevertheless, some least developed countries (LDCs) have made significant progress. For example, implementation rates in Bangladesh and Cambodia were above the regional average. Research shows that South Korea leads East and Northeast Asia in establishing legal frameworks and AI governance. Kazakhstan leads North and Central Asia. Singapore leads Southeast Asia. And Bangladesh leads South Asia and Southwest Asia. This development is encouraging for Bangladesh.
According to the report, nearly 50% of Asia-Pacific countries have started working on AI systems to automate customs procedures, but the implementation rate of concrete measures is much lower.
Overall, these countries showed that AI is most used for fraud/smuggling detection and cargo inspection/image analysis. Customs is also far ahead of other government agencies in applying AI to trade facilitation.
There are several barriers to using AI in trade facilitation. According to the survey results, the biggest barriers in the developing Asia-Pacific region are a lack of AI and machine learning expertise and skills, as well as coordination challenges and high costs.
The findings and analysis presented in the above report have important implications for policy makers in Bangladesh and other developing countries. Trade facilitation aims to reduce costs, but ongoing geopolitical conflicts and tensions are increasing trade costs in many ways. This makes it more difficult for countries to contain costs and advance trade facilitation measures. The use of AI can help countries reduce trade costs.
asjadulk@gmail.com
