The artificial intelligence boom is being fueled by an intricate circular financing loop that has drawn billions of dollars from foreign investors, according to a review of corporate filings and market data. The arrangement, in which AI companies, their suppliers, and their customers invest in one another, is inflating valuations and raising comparisons to the debt structures that preceded the 2008 financial crisis.
How the loop works
At its core, the loop involves a handful of large AI firms and their close partners. Chipmakers invest in AI startups that then buy chips. Cloud providers strike deals with AI model developers, who in turn commit to spending billions on cloud services. The same money, in effect, cycles through multiple balance sheets, each time recorded as revenue or growth.
Foreign capital has become a critical part of the cycle. Sovereign wealth funds and overseas institutional investors have poured billions into AI-focused funds and direct stakes over the past 18 months. That money helps sustain the loop, but it also creates a web of cross-holdings that can obscure where the ultimate risk lies.
Debt levels draw comparisons to 2008
The rising debt across the AI sector is what worries some market participants. Companies that a few years ago carried little leverage are now issuing bonds and taking on loans to fund data-center construction, chip purchases, and acquisitions. The total debt load has grown quickly, though exact figures are hard to pin down because much of it sits in private vehicles and off-balance-sheet arrangements.
Those dynamics echo the pre-2008 period, when complex credit structures and interconnected counterparties turned a downturn in one corner of the market into a global freeze. The parallel isn't exact — today's AI firms are generally more profitable than the subprime borrowers of that era — but the opacity and the circularity are similar enough to make regulators and risk managers uneasy.
Foreign money adds a layer of complexity
Cross-border investment brings its own set of complications. When foreign entities hold significant stakes in AI companies that also rely on each other for revenue, a pullback from one investor can ripple through the entire chain. Currency fluctuations, changing tax rules, and geopolitical tensions can all prompt foreign capital to retreat faster than domestic investors might.
For portfolio managers, the challenge is that traditional diversification may not work as intended. If AI stocks, AI bonds, and the private funds that hold both are all exposed to the same underlying cash flows, a sell-off in one asset class can drag down the others. Building a resilient portfolio in this environment requires looking through the circular ties, not just at headline allocations.
What could go wrong
The most immediate risk is a funding squeeze. If foreign investors slow their commitments, AI companies that depend on that capital to cover operating losses could face cash shortages. That, in turn, would pressure their suppliers and partners, who are often the same firms propping each other up. A single high-profile default or missed payment could trigger a broader reassessment of the sector's creditworthiness.
Market destabilization wouldn't necessarily require a collapse. Even a moderate repricing of AI-related debt could force leveraged holders to sell other assets, spreading losses to unrelated markets. Investor strategies built on the assumption that AI growth is a one-way bet would need to be rethought quickly.
For now, the loop keeps spinning, and the billions keep flowing. But the connections that make the AI sector seem unstoppable are the same ones that could make a downturn harder to contain. The next test may come when a major funding round or debt maturity forces the participants to show whether the money circulating among them is as solid as it looks.
