Executive Summary
A new finding from the Bank for International Settlements puts a sharper edge on a question that has hovered over the AI boom: how much of the sector’s momentum is being reinforced from within. According to the available source information, 28.7% of AI investment deals involve circular relationships in which AI firms invest in each other. The BIS frames that pattern as a source of systemic macroeconomic risk and supply-chain opacity.
That does not, by itself, invalidate the AI buildout or prove that demand is overstated. But it does suggest that a meaningful share of capital flows inside the sector may be harder to interpret than conventional outside investment. When companies in the same industry help finance one another, investment activity can become less useful as a clean signal of independent market conviction.
For TechPowerAsia readers, the importance is broader than venture finance. AI is now a capital-intensive industrial system spanning chips, servers, memory, networking, cloud infrastructure, and power-hungry data center expansion. Any development that blurs the visibility of underlying demand matters for Asia, because Asia remains central to the manufacturing and supply chains behind global AI deployment.
The BIS report is global rather than Asia-specific. Still, the strategic implication is clear: if AI capital is circulating inside a more interdependent corporate network than many observers assumed, then supply-chain transparency, financial resilience, and demand verification all become more important.
Watch the Short Brief
Watch this short visual briefing for the key strategic implications behind the story.
Key Developments
According to the BIS report, 28.7% of AI investment deals involve circular relationships among AI firms. In practical terms, that means companies in the sector are not only building products and buying infrastructure, but also acting as financial backers of one another.
The BIS characterizes this pattern as a potential source of systemic macroeconomic risk. That framing matters. A central bank institution is not simply describing a niche venture-capital trend; it is identifying a structure that could complicate how policymakers, investors, and industry participants assess the stability of a fast-growing technology sector.
The report also points to supply-chain opacity. Based on the available source information, the concern is not merely that AI financing is active, but that intertwined capital relationships may make it harder to understand who depends on whom, where risk is concentrated, and how stress might travel across the ecosystem.
What the current source material does not establish is equally important. It does not identify specific companies, regions, or transaction types involved in these circular arrangements. It also does not, in the materials provided here, quantify which subsegments of AI are most exposed. The confirmed takeaway is narrower and more structural: circular investment is present at a notable level, and the BIS considers it significant enough to raise macro-risk and transparency concerns.
Strategic Analysis
The main strategic issue is signal quality. In most technology cycles, capital flows serve as an important external validation mechanism. If independent investors continue to fund an industry at scale, markets often read that as evidence of confidence in future demand, margins, and adoption. Circular investment complicates that reading.
When one AI firm invests in another AI firm, the capital itself may still be rational and economically justified. Strategic investment is a familiar part of the technology industry. Companies often back adjacent platforms, infrastructure partners, or emerging capabilities that could strengthen their own positions. The problem is not that such investment exists. The problem is that, at a certain scale, it can blur the line between independent demand validation and internally reinforced momentum.
That distinction matters more in AI than in many previous software cycles because the sector’s growth increasingly rests on heavy physical investment. AI expansion is tied to semiconductor capacity, advanced packaging, high-bandwidth memory, server assembly, optical interconnects, power systems, and data center construction. Those are long-lead, capital-intensive systems. If financial signals around AI become harder to interpret, then downstream planning across the hardware stack may also become harder to calibrate.
This is where the Asia relevance comes into focus. The BIS finding is global, but the world’s AI supply chain is not geographically neutral. Asia plays a central role in semiconductor fabrication, packaging, components, electronics manufacturing, and infrastructure deployment. That means any opacity in AI capital formation can carry indirect consequences for Asian firms even when the original investment relationships are elsewhere.
One implication is around demand verification. Suppliers, manufacturers, and capital allocators often rely on multiple indicators when judging the durability of an upcycle: customer orders, capacity reservations, equipment spending, financing activity, and broader market sentiment. If a sizable share of AI investment deals reflects circular relationships, then financing data alone may offer a less reliable guide to underlying end demand than headline figures suggest.
Another implication is around correlation risk. Circular investment does not automatically create instability, but it may increase interdependence within the sector. If companies are connected not only through commercial relationships but also through ownership ties, stress in one part of the AI ecosystem could become more difficult to isolate. That does not mean a disruption would necessarily spread widely. It does mean the network could be more tightly coupled than investors assume from operating metrics alone.
The supply-chain opacity point deserves particular attention. In AI, visibility is already uneven. Many critical relationships are mediated through contract manufacturing, cloud platforms, infrastructure partnerships, and multi-layered procurement channels. If cross-investment adds another layer of complexity, outside observers may find it harder to determine whether growth is being driven primarily by diversified customer demand, strategic ecosystem support, or a combination of both.
For policymakers, that makes the BIS warning more than an abstract financial observation. AI is increasingly treated as strategic infrastructure. Governments across Asia are making decisions about energy allocation, data center policy, semiconductor incentives, digital sovereignty, and national AI competitiveness. Those decisions depend on clear market signals. If parts of the investment landscape are circular, the risk is that public and private decision-makers may overread the independence of those signals.
None of this should be overstated. The BIS finding does not prove that AI investment activity is unsound, nor does it show that circular deals are inherently destabilizing in every case. Strategic cross-investment can support ecosystem development, accelerate commercialization, and align incentives across a young industry. But the BIS appears to be highlighting a threshold issue: when this behavior becomes common enough, it becomes a financial-stability question as well as a corporate-strategy question.
That is especially relevant for Asia-focused technology intelligence. The region’s exposure to AI is not limited to model developers or software platforms. Asia’s importance lies in the industrial base that makes AI scale possible. Any distortion in how the market interprets AI demand, capital formation, or supply-chain resilience could affect investment pacing, procurement expectations, and infrastructure planning across the region.
Investor Takeaway
The BIS finding is best read as a structural risk signal, not as a verdict on the AI sector. The key message is not that circular investment is inherently problematic, but that it can reduce transparency at a time when markets are trying to assess the true durability of AI spending.
For investors and industry decision-makers, the first priority is to watch disclosure quality. The more clearly companies describe the nature of strategic investments, ownership links, and ecosystem financing relationships, the easier it becomes to distinguish independent demand from intra-sector support. Limited disclosure would leave more room for misreading growth signals.
Second, investors should pay closer attention to the difference between capital activity and operating demand. In an AI market defined by aggressive infrastructure spending, headline funding momentum can look persuasive on its own. But the more important question is whether demand is broadening across end users, workloads, and business models, rather than being reinforced mainly within a concentrated network of sector participants.
Third, Asia-focused observers should monitor whether future reporting clarifies regional exposure. The current source material does not identify Asia-specific participants or vulnerabilities. Even so, because Asian supply chains underpin much of the global AI hardware stack, any persistent opacity in AI capital flows could eventually influence production planning, supplier expectations, and investment timing across the region.
Finally, the BIS report suggests that AI should be evaluated not only as a technology story, but also as a capital-structure story. That shift matters. In the current cycle, the strategic question is no longer just who has the best models or the most compute. It is also who is financing whom, how visible those relationships are, and whether the market’s demand signals remain clean enough to support sound industrial and financial decision-making.
For TechPowerAsia readers, that makes circular AI investment worth tracking closely. In a sector where semiconductors, infrastructure, and capital flows are deeply intertwined, opacity is not a side issue. It is part of the core risk map.