Key Notes
- Georgieva puts AI circular financing at just over $1 trillion, with about two-thirds funded by debt.
- She warns that an AI investment retreat could weaken growth and threaten financial stability.
- She describes the recent market adjustment as orderly rather than a developing systemic crisis.
International Monetary Fund Managing Director Kristalina Georgieva warned Monday that the global economy’s reliance on artificial intelligence investment could become a financial-stability problem if confidence in the sector breaks. Circular financing among AI companies has reached just over $1 trillion, with about two-thirds funded through debt, she said.
Speaking to CNBC on the sidelines of the ASEAN summit in the Philippines, Georgieva described an economy “standing on one leg” because of its dependence on AI. “If this leg wobbles, that can lead to financial stability troubles,” she said.
Her assessment of the recent market selloff was more measured. Georgieva characterized the adjustment as orderly and considered it unlikely to develop into a disruption that would undermine the world economy. Her concern centers on the speed at which financing ties are growing and what a sudden investment retreat could expose.
How the Financing Loop Works
Circular financing arises when a company invests in, or helps finance, another business that then uses the money to buy its products or services. In AI, the same firms can become one another’s investors, suppliers and customers.
For example, financing can help a cloud operator acquire accelerators and build a data center. The supplier earns hardware revenue as equipment is delivered, while the operator expects to repay its funding from future sales of computing capacity. The strength of that arrangement depends on demand from the customers who ultimately use the infrastructure.
An October IMF staff note describes how AI investment is moving beyond internally generated cash toward external funding and more interconnected structures. It identifies private credit as a growing link among banks, nonbank lenders and AI borrowers.
Nvidia (Nasdaq: NVDA) illustrates the overlapping roles. Its customer-financing arrangements include investments, cloud-capacity purchases and contingent guarantees. Those contracts expose the chipmaker to different risks alongside the revenue it earns from selling equipment.
Debt Can Spread the Shock
A slowdown in AI demand can travel through that network. A cloud operator receiving less cash from customers may cut equipment orders or struggle to service borrowing. An investor or supplier that guaranteed some obligations could then face a payment claim as well as weaker sales.
The IMF’s October financial-stability report urged supervisors to assess potential systemwide losses from AI exposures, including the possibility that circular financing amplifies distress. Where vulnerabilities are material, it pointed to concentration limits and capital buffers among the tools authorities could use.
Georgieva’s warning puts the focus on the cash generated after the infrastructure is built. Debt remains payable even if a data center fills more slowly than expected. A drop in bookings can therefore reach lenders before a project has time to find replacement customers.