POSCO International said on July 27 that it had successfully completed a proof of concept with LG CNS to test trade finance infrastructure using blockchain and artificial intelligence. The verification process ended on July 23 and was designed to assess how digital technologies could raise operational efficiency, automate trade-related processes, and strengthen risk management across POSCO International’s expanding global business network.

The initiative was created to support POSCO International’s growing international operations by improving how the company manages transactions and settlements involving overseas subsidiaries. As its global footprint continues to increase, the project also fits into the company’s wider digital transformation strategy for trade operations, where documentation, settlement coordination, and counterparty verification often involve multiple internal teams, banks, customers, and logistics-related participants.

Pilot deployment planned after digital trade finance verification

The proof of concept validated the use of blockchain-based shared ledgers, AI-driven automation, and tokenized real-world assets (RWAs) as tools for modernizing trade finance and improving efficiency in global trading operations.

During the project, POSCO International and LG CNS evaluated three core technology areas: shared ledgers based on blockchain for improved transaction visibility, tokenization of trade receivables as real-world assets, and AI agents intended to automate document-heavy trade workflows.

The verification examined whether transaction management, settlement work, and document review tasks generated in international trade could be digitally integrated and automated. POSCO International provided its global trade environment and real transaction data to create conditions that reflected practical business operations. LG CNS handled system architecture design and validated the blockchain and AI technologies used in the test.

A central part of the project was the implementation of a blockchain-based shared ledger that would allow headquarters, overseas subsidiaries, and business partners to view the same transaction information in real time. Before this approach, transaction data was often managed separately by regional offices or subsidiaries, which required repeated verification during contract execution and settlement procedures. The shared-ledger model showed potential to reduce inconsistencies in transaction records, improve collaboration, and lower operational risks linked to information discrepancies.

The companies also evaluated the tokenization of trade receivables as real-world assets. Under the proposed structure, receivables created through actual commercial transactions could be converted into digital assets that can be transferred, traded, managed, and settled using blockchain infrastructure. This part of the verification was conducted using Injective, a blockchain network designed for enterprise financial applications.

The project further assessed whether enterprise-grade compliance requirements could be built directly into the blockchain protocol. The review covered permission-based asset management, Know Your Customer (KYC) procedures, Anti-Money Laundering (AML) compliance, investor eligibility checks, and restrictions on asset transfers. The purpose of these findings was to determine whether blockchain technology could support business-to-business receivables management while meeting regulatory and privacy requirements.

That compliance layer is a key practical issue for applying blockchain to corporate finance rather than consumer-facing digital assets. For trade receivables, companies must be able to control who can hold or transfer an asset, verify participants, and maintain auditable transaction records while protecting commercially sensitive information.

The AI component of the proof of concept focused on automating reviews of letters of credit (LCs) and other trade documents. AI agents were tested for their ability to detect document errors and support compliance checks before transactions advanced further through the trade process.

According to the results, the AI agent demonstrated that it could automate document reviews for letters of credit and trade documentation, helping identify errors at an earlier stage and reducing inconsistencies caused by differences in staff expertise.

International trade documentation often requires specialized knowledge because regulatory obligations and contractual terms can vary by jurisdiction, counterparty, and transaction structure. Even minor errors, such as typographical mistakes or missing contractual clauses, may lead to delayed settlements or rejected payments. POSCO International expects AI-supported preliminary reviews to improve document quality and reduce operational differences among its international offices.

After the verification was completed, a company representative said the project had demonstrated the practical applicability of blockchain and AI technologies by using real-world trade data and operational processes. The representative also said the company plans to develop a pilot deployment strategy during the second half of the year, with a focus on business areas where measurable operational improvements were confirmed.

Following the successful proof of concept, POSCO International intends to prepare a pilot deployment later this year, concentrating on trade finance functions where blockchain and AI produced the strongest operational benefits. The next stage will indicate which parts of the verified infrastructure can move from a controlled test into operational use across selected trade finance processes.