Dominican Republic Customs to Add AI to X-Ray Inspections

The General Directorate of Customs (DGA) is preparing to introduce artificial intelligence into the analysis of X-ray images generated during non-intrusive inspections of cargo, expanding the use of technology in the Dominican Republic’s customs risk-management system.

DGA Director General Nelson Arroyo said the system will allow X-ray images to be evaluated simultaneously by an advanced AI model capable of supporting the identification of inconsistencies, unusual patterns and elements that may require specialized inspection. The announcement was made during the Industrial Meeting “Customs as a Strategic Ally for Productive Development,” organized by the Association of Industries of the Dominican Republic (AIRD).

The initiative would add an automated analytical layer to inspections that currently depend heavily on trained customs personnel interpreting scanner images. The objective is not to replace those officers, but to give them an additional source of information when determining whether a shipment warrants closer examination.

AI Would Connect X-Ray Analysis With Customs Risk Management

The planned system forms part of a broader shift by the DGA toward data-driven risk management. Arroyo said artificial intelligence will also be incorporated into the agency’s risk-management engine to identify atypical patterns and improve the profiling of cargo before or during customs processing.

The technology is expected to help identify anomalies associated not only with the physical contents of shipments but also with the determination of customs tax obligations. That means the potential application goes beyond detecting concealed goods: the same broader risk-management approach can be used to identify transactions or declarations that merit additional scrutiny.

This direction is consistent with the DGA’s 2026 operational plan, which identifies efficient scanning and intelligent customs risk management as institutional priorities. The plan sets a target for 90% of container scans to be completed in less than five minutes and calls for detailed physical inspections when discrepancies are identified in X-ray images.

The DGA’s recent performance figures also show why the agency is placing greater emphasis on targeted intervention. During the first seven months of 2026, its risk-management system generated 14,774 alerts in primary customs zones. According to figures presented by Arroyo, interventions resulting from those alerts generated more than RD$2.63 billion in additional revenue, while the share of declarations receiving alerts fell from 28.6% in 2025 to 25% during January-July 2026.

Central Monitoring Is Another Part of the Plan

Arroyo also said the DGA wants to monitor from its central headquarters, in real time, the work performed by operators of X-ray machines installed at the country’s ports. Centralized monitoring would give the institution greater visibility over how inspections are being conducted and could allow supervisors to identify discrepancies or operational issues without relying exclusively on reports from individual facilities.

The combination of centralized monitoring, X-ray analysis and automated risk profiling would give the customs administration a more integrated view of cargo movements. For importers and exporters, the intended balance is significant: high-risk shipments can receive greater scrutiny while compliant operations can move through the system with fewer unnecessary interventions.

The World Customs Organization (WCO) describes non-intrusive inspection as technology that allows customs authorities to examine cargo without opening containers, vehicles or packages. The organization considers these systems an important tool for controlling high-risk cargo while minimizing disruption to legitimate trade.

The Technology Is Already Being Tested by Customs Administrations Worldwide

The Dominican Republic is not developing this approach in isolation. The WCO reported in March 2026 that customs administrations are increasingly exploring AI and machine learning to analyze the large volumes of information generated by inspections and risk-management systems.

India, for example, has developed an AI-based X-ray image analytics system designed to identify anomalies, concealed goods and possible misdeclarations. Its customs administration combines AI-generated results with the judgment of human officers, who ultimately determine whether an image should be treated as suspicious. The WCO says the system has been designed to improve detection while reducing inspection times and strengthening risk-based selection.

China Customs has taken the technology considerably further. According to the WCO, its AI-based image-analysis system has been deployed on hundreds of non-intrusive inspection devices and can identify prohibited items, classify thousands of types of goods, compare image results with customs declarations and detect concealed compartments in vehicles and trains. The system reportedly generated alarms that contributed to the detection of more than 20,000 smuggling attempts since 2022.

These international examples illustrate the potential capabilities of the technology, but they should not be interpreted as evidence that the Dominican system will initially offer the same functions. The DGA has announced the planned use of AI for X-ray analysis and risk management, but the specific models, training data, detection categories, accuracy levels and implementation timetable have not been publicly detailed in the information currently available.

Dominican Customs Is Expanding Its Digital Infrastructure

The X-ray initiative is part of a wider modernization program at the DGA. The agency has been expanding digital services, automated risk analysis and mechanisms designed to differentiate between compliant operators and shipments presenting higher risks.

One of the most established components is the Authorized Economic Operator (OEA) program. As of August 2026, more than 700 companies were certified under the standard and simplified OEA schemes, and those companies accounted for more than 40% of Dominican imports and a similar share of customs collections, according to figures presented by Arroyo. More than 96% of their operations were reportedly processed through lower-intervention channels.

The broader strategy is also consistent with recommendations and observations from the International Monetary Fund. Its 2025 assessment of the Dominican Republic highlighted the continued modernization of customs administration, including greater use of technology, digitalization, interoperability and data analytics to streamline procedures, strengthen controls and detect anomalies or fraud.

Faster Trade Processing Remains a Key Objective

The push toward automated inspection is taking place as the Dominican Republic seeks to strengthen its position as a regional logistics and trade hub. The government has increasingly linked customs modernization with shorter clearance times, improved supply-chain performance and greater predictability for companies engaged in international trade.

The AIRD has similarly argued that customs efficiency affects much more than government revenue. During the August industrial meeting, AIRD President Julio Virgilio Brache said a modern customs administration is part of the country’s competitive infrastructure because its performance influences supply chains, exports and companies’ ability to produce, invest and create jobs.

The DGA is also developing a logistics activity and efficiency monitor intended to identify the causes of delays affecting foreign-trade operations and support measures to reduce logistical time and costs.

For the Dominican Republic, the significance of the AI project therefore extends beyond the scanners themselves. If implemented effectively, automated image analysis could allow customs officers to concentrate more attention on shipments that present genuine risk while reducing unnecessary disruption to legitimate cargo. The success of the system will ultimately depend on the quality of its data, the accuracy of its models, the integration of different customs databases and the ability of human officers to validate the alerts generated by the technology.