Artificial intelligence (AI) has started to protect South Korea's borders. By taking a photo of the ingredient list of imported food and pharmaceuticals with a smartphone, AI can determine whether the items are harmful. This process, which previously took over five minutes, has been reduced to about five seconds.
AI is also being used to detect reselling and tax evasion disguised as overseas direct purchases, as well as to track complex trade and foreign exchange crimes.
At the center of this transformation is Lee Jong-wook, the Commissioner of the Korea Customs Service.
The hallmark of Lee's AI policy is clear. It does not merely involve importing a massive AI system from outside. Instead, field officials, who have the best understanding of customs operations, identify problems and create solutions using AI.
In short, it is an 'AI Transformation (AX) starting from the field.'
The Korea Customs Service has been training hybrid talents with expertise in customs operations and big data and AI analysis since 2017. In 2026, it selected 50 AI analysts from customs offices nationwide to form the AX promotion team.
During the first half of 2026, the 'Korea Customs Service AX Challenge' featured 22 cases led by field AI analysts, with eight advancing to the finals. The Customs Service has expanded its previous 'AI and Big Data Awards' to hold the AX Challenge twice a year starting this year.
AI Government's Answers Are Found in the Field
As the era of AI revolution unfolds, government agencies and public institutions are competing to adopt AI. However, simply establishing an AI system does not equate to innovation.
Even if hundreds of millions are spent on creating an AI platform, it is meaningless if it is not used in the field. Conversely, if even a small AI program reduces the workload of officials and provides better services to the public, that is true AI innovation.
This is the point that Lee Jong-wook emphasizes. He stated during the first half of 2026's Korea Customs Service AX Challenge, "To realize an AI government, each customs official at the citizen interface must be able to utilize technology to connect policy and field administration execution."
Government AI innovation does not occur solely through directives from ministers or commissioners. Field officials must understand AI and apply it to their work. The essence of Lee's AI administration lies here.
It may be quicker to teach customs experts AI than to teach AI experts about customs operations.
AI Solves Tasks That Previously Took Five Minutes in Just Five Seconds
The best example of the Korea Customs Service's AI innovation is the 'model that blocks illegal food and pharmaceuticals from abroad with a single photo.' The food and pharmaceuticals imported from overseas are directly related to public health. Customs officials must verify product names and ingredient lists to determine whether they are harmful.
Previously, this verification took over five minutes. With AI, the situation has changed. By taking a photo of the product name and ingredient list with a smartphone, AI analyzes the information and checks in real-time whether the items are harmful. The task that once took over five minutes now takes about five seconds. This model was developed by Lee Byung-seok, a customs officer at Gwangju Customs, and won the grand prize at the first half of the 2026 Korea Customs Service AX Challenge.
AI innovation does not necessarily have to be a massive project. It starts with identifying everyday inconveniences in the field and solving them with AI. If small innovations spread to thousands of officials and benefit millions of citizens, that is the AI transformation the government should pursue.
AI Detects Trade and Foreign Exchange Crimes and Tax Evasion
AI's role extends beyond reducing work hours. It is also utilized to detect increasingly sophisticated crimes. At Seoul Customs, an AI-based trade and foreign exchange crime tracking system has been developed.
In Incheon Customs, a model was created to identify reselling and tax evasion by individuals who import goods as if for personal use and then sell them domestically, using AI address recognition. In the past, customs officials heavily relied on experience and intuition.
However, there are limits to how much data and transactions a person can analyze individually. AI can quickly analyze vast amounts of data and identify anomalies and patterns that are difficult for humans to detect.
By combining the experience of customs officials with AI's data analysis capabilities, a new border management system is being established.
Why the Customs Service Focuses on '50 AI Analysts'
What is most important in the AI transformation? Many people talk about technology, saying that GPUs, data centers, and large language models are necessary. While this is not incorrect, the real change in an organization ultimately comes from people. The Customs Service has recognized this.
Since 2017, it has been training talents with expertise in customs operations and AI and big data analysis. In 2026, it selected 50 AI analysts from customs offices nationwide.
These individuals are not just AI experts. They are hybrid talents who understand customs administration and can utilize AI. This has significant implications for organizational innovation.
In the AI era, it is not necessary to make every official an AI developer. Instead, it is essential to cultivate key personnel who understand AI within each organization and solve field problems centered around them. If 50 AI analysts become 100 or 500, the way the Customs Service operates as a large government organization could change.
Innovation from the Bottom Up, Not Top Down
Another noteworthy aspect of the Customs Service's AX policy is the direction of innovation. Traditional government innovation is mostly top-down, where policies are created at the headquarters and directed to local organizations.
However, the Customs Service's AX Challenge is different. It identifies problems in the field, creates solutions using AI, discovers best practices, and, once verified, spreads them throughout the organization. This is 'AI innovation rising from the bottom up.'
The advantages of this approach are clear. By solving problems that field officials know best, the likelihood of practical application in actual work increases. Additionally, the experience of knowing that one's ideas can change the organization encourages voluntary AI utilization among officials.
The Customs Service plans to continue discovering and sharing best practices with the public following the first half of the 2026 AX Challenge.
The Role of the Customs Service Becomes More Important in the AI Era
The global economy is rapidly digitalizing. E-commerce and overseas direct purchases are surging, and global supply chains are becoming increasingly complex. The methods of smuggling drugs and illegal goods are also becoming more sophisticated. The era of managing borders solely with human power is coming to an end.
There is a need to analyze countless items, transactions, companies, and movements in real-time. This is why AI is necessary. The Customs Service possesses far more data than the public realizes.
Export and import data, company information, logistics data, traveler information, and other vast amounts of data are accumulated during the customs administration process.
When AI and data are combined, the Customs Service can evolve from merely being an agency that collects taxes and enforces illegal goods to an 'AI economic security agency' that predicts changes in global supply chains, supports corporate exports, and detects economic security risks in advance.
Can Lee Jong-wook's Experiment Change the South Korean Government?
The South Korean government employs countless officials who perform numerous repetitive tasks daily. They write reports, analyze complaints, search for data, and review various documents.
How many of these tasks can be solved with AI? The Customs Service's experiment raises an important question. While it is necessary for the government to build a massive AI system, is it not more crucial for each official to utilize AI in their work?
Recently, the Customs Service was awarded the 'Presidential Award for Contributions to AI Government Development' for recognizing the achievements of its AX promotion, which combines short-term tasks with long-term strategies.
Lee Jong-wook's AI policy focuses on changes in the field rather than flashy slogans. It reduces tasks that took five minutes to five seconds, uncovers traces of crimes that humans might miss, and applies the ideas of field officials to actual administration.
The national AI transformation does not only begin in grand places. It starts with solving small problems in the field using AI. This is why the Customs Service's experiment, led by Lee Jong-wook, is gaining attention.
The South Korean government's AI transformation may find its answers, as Lee Jong-wook emphasizes, in the field closest to the citizens.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.