INDEPENDENT JOURNAL

A clearer view of what matters

An independent news site covering AI, technology, economy, society, science, health, sports, culture and world affairs.

AI · 2026-11-01

KB Kookmin Bank Leads 12.6 Trillion Won AI Infrastructure Financing: What GPU Financing Means and What Comes Next

KB Kookmin Bank has been selected as the lead arranger for Brookfield’s AI computing infrastructure project, worth about 12.6 trillion won. Around 10.7 trillion won is expected to be raised through a syndicated loan involving domestic banks and securities firms.

KB Kookmin Bank Leads 12.6 Trillion Won AI Infrastructure Financing: What GPU Financing Means and What Comes Next
AI-generated editorial illustration
As AI services grow, the computing equipment and infrastructure that support them require enormous amounts of capital. With KB Kookmin Bank securing the lead arranging role in Brookfield’s large-scale AI computing infrastructure project, attention is turning to whether financing for technology assets such as GPUs will begin to take hold in Korea as well.

The project is valued at about 12.6 trillion won. Of that amount, roughly 10.7 trillion won is planned to be raised through a syndicated loan involving major domestic banks and securities firms. A syndicated loan is a structure in which multiple financial institutions share the loan, helping to spread out the funding burden and risk that would otherwise fall on a single institution.

According to License News, KB Kookmin Bank was chosen as the lead arranger for the deal and will oversee financing work, including designing the financial structure and coordinating key lending terms. The project is structured so that a special purpose company (SPC) established by Brookfield acquires GPUs and other AI computing equipment and related infrastructure, then supplies them to Naver Cloud over the long term. The funds raised are expected to be used to purchase GPUs and related equipment and to build out the infrastructure.

However, securing lead arranging rights does not mean the loans have already been executed. Based on the report, the current stage is one of arranging financing, while the actual loan size and each institution’s participation amount may be determined later during the financing agreement and disbursement process. For that reason, it is important not to confuse the overall project value with the finalized amount of loan funding.

What makes this transaction noteworthy is that it applies traditional infrastructure finance methods to AI infrastructure investment. The idea is to raise funds based on the cash flow expected from a long-term supply contract. The bank described this as Korea’s first GPU financing and the largest syndicated loan arrangement in the country’s history. That wording should be understood in light of KB Kookmin Bank’s explanation as reported.

The key issue going forward will be how the loan agreement and actual funding are carried out. GPUs are essential for AI computing, but because technology changes so quickly, their performance and market value can shift over time. The stability of the long-term supply contract, the residual value of the equipment, and how the debt repayment structure is managed are all important risk factors as well. This financing arrangement may become both a new example of providing capital for the growth of the AI industry and a case study in what kind of risk management is needed for large-scale financing backed by technology assets to remain sustainable.
← Back to articles