Bank of Korea Warns: Mounting Vulnerabilities in the AI Sector from Heavy Reliance on External Financing and Intensifying Circular Finance as Long-Term Interest Rates Rise

On October 27 that, according to Yonhap News Agency, Section Chief Kim Yun-kyung and Investigator Wang Jae-woon from the Bank of Korea's Research Bureau stated in the report "Characteristics and Risk Assessment of Financing in the AI Industry" that the recent expansion of funding competition within the AI sector has heightened reliance on external capital and cyclical financing, prompting growing warnings about AI-induced financial risks. The report notes that hyperscale data center operators have significantly broadened their external financing this year, primarily through long-term corporate bonds, while AI-focused cloud companies have heavily relied on GPU-collateralized loans and convertible bonds. AI firms are also actively leveraging off-balance-sheet financing structures such as special purpose vehicles, joint ventures, and leasing. Furthermore, the scale of circular transactions involving forward sales agreements between hyperscale operators and AI chipmakers via supply chain investments continues to grow.

The report evaluates that while dominant leading companies maintain robust financial positions, reducing the likelihood of AI-related financial risks materializing, underlying vulnerabilities are steadily accumulating amid rising long-term interest rates. Key risk factors include delayed AI commercialization and the spillover effects of cyclical financing, necessitating close scrutiny of AI cloud providers characterized by inadequate financial buffers and poor information transparency. Highlighting that South Korea is the greatest beneficiary of the AI boom, the report stresses that due to its extensive ties to global supply chains, the transmission intensity of real-economy and financial shocks triggered by AI could exceed historical levels. Continuous assessment of building vulnerabilities and risk triggers remains essential.