A new CRS report warns circular finance among AI companies could create systemic financial risk. Read the full breakdown of what Congress is weighing.
Financial arrangements where artificial intelligence (AI) companies invest in other businesses that then purchase their products or services, known as circular financing, could expose the broader economy to financial instability. This warning comes from a new Congressional Research Service (CRS) report. The practice has surged as technology companies channel billions into AI infrastructure, with analysts projecting U.S. AI-related infrastructure investment to be around $500 billion this year and a staggering $10.3 trillion between 2025 and 2032. These arrangements are prompting Congress to scrutinize whether the rapid expansion of the AI industry could trigger economic or financial shocks, and how circular financing might intensify such disruptions. While the report does not definitively declare an AI investment bubble or an imminent financial crisis due to circular financing, it highlights significant concerns regarding potential systemic risks.
Circular financing is characterized by a company providing financial support to another business with the understanding that the recipient will in turn purchase its products or services. This practice is increasingly prevalent among AI developers, chipmakers, and cloud computing providers. A study by the Bank for International Settlements revealed that nearly 28.7% of investments made by AI companies between 2021 and 2025, based on deal value, were directed towards other AI companies. Furthermore, 55.2% of the total investment flowing into AI companies originated from other AI businesses. The report outlines three main types of circular financing: investor-supplier relationships (where a company finances a customer's product purchase), investor-purchaser relationships (where a customer invests in its supplier), and reciprocal arrangements involving mutual financial commitments. Notable examples include Nvidia's announced intention to invest up to $100 billion in OpenAI, which aimed to develop computing infrastructure powered by Nvidia technology (though this plan later reportedly stalled). Another instance is Amazon's additional $5 billion investment in Anthropic, with a potential for $20 billion more, contingent on Anthropic committing to spend $100 billion over 10 years on Amazon's chips and cloud computing services. These arrangements are seen as an evolution of vendor financing, a long-standing practice where suppliers aid customers in financing purchases, offering capital and unique risk assessment capabilities. While this financing is crucial for many AI startups, which often operate at a deficit and rely on external investment to scale, it complicates the assessment of whether demand for AI products truly reflects market growth or is merely sustained by supplier-backed spending. The report cautions that complex financial ties can obscure interdependencies among AI companies, potentially downplaying their exposure to industry-wide financial issues, drawing parallels to the circular financing observed between telecommunications equipment suppliers and customers during the dot-com boom, which led to failures when funding dried up.
The expanding financial connections within the AI sector could allow problems at one company to quickly spread to its suppliers, investors, and customers. If major banks or other financial institutions become heavily invested in these arrangements, disruptions within the AI industry could trigger wider repercussions across global financial markets. The International Monetary Fund’s October Global Financial Stability Report highlighted a potential correction in AI-related investment as a significant concern for financial market participants. However, the degree of systemic risk is partly contingent on the funding sources; if AI investments are financed through companies' own operating revenue rather than through borrowing from banks, the immediate exposure to the broader financial system may be more contained. For Congress, the central question remains whether circular financing genuinely facilitates the development of essential AI infrastructure or if it inadvertently creates financial dependencies that could exacerbate the impact of an industry downturn, posing a complex challenge for future policy decisions.