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
According to Yonhap News Agency, Section Chief Kim Yun-kyung and Investigator Wang Jae-kun of the Bank of Korea’s Investigation Bureau noted in their October 27 report, "Major Characteristics and Risk Assessment of Funding in the AI Industry," that intensifying fundraising competition in the AI sector recently has heightened dependence on external funding and expanded circular finance, fueling growing warnings about financial risks triggered by AI. The report notes that this year, hyperscale data center operators have substantially broadened external financing primarily through long-term corporate bonds, while AI-focused cloud companies have increasingly turned to GPU-backed credit and convertible bonds. AI firms are also actively leveraging off-balance-sheet financing instruments such as special purpose vehicles, joint ventures, and leasing. Meanwhile, the volume of circular transactions between hyperscale firms and AI chipmakers for future sales, facilitated by supply chain investments, continues to grow. The report assesses that while the financial health of major leading enterprises remains solid and the likelihood of AI-related financial risks materializing is currently low, vulnerabilities are steadily building amid rising long-term interest rates. Primary risk factors include delays in AI monetization and the propagation of shocks stemming from circular finance. Particular attention should be directed toward the potential risks posed by AI cloud companies that exhibit inadequate financial buffers and low information transparency. The report emphasizes that as South Korea stands to benefit the most from the AI boom, its deep integration into global supply chains could amplify the transmission intensity of real-economy and financial shocks triggered by AI compared to previous cycles. It is essential to continuously monitor the accumulation of potential vulnerabilities and the likelihood of risks materializing.
