Bloomberg Analysis Warns of Risks in AI's Circular Funding Models

A new analysis by Bloomberg has raised concerns about the financial interdependence within the AI sector, particularly through circular funding models that involve key players like Nvidia, OpenAI, and Anthropic. These models allow companies to invest in one another, creating a cycle where investments from one firm are used to purchase products from another, thereby inflating revenues and valuations.

According to Bloomberg, over 55% of incoming investment to AI firms between 2021 and 2025 originated from other AI companies. This interconnectedness raises alarms about potential financial instability reminiscent of the 2007-08 financial crisis. The report emphasizes that while these funding structures may accelerate innovation, they also amplify financial imbalances.

Key findings from the Bank for International Settlements indicate that 28.7% of outbound investments by AI firms were directed towards other AI firms, suggesting a complex web of dependencies. Notably, Nvidia's financial ties with CoreWeave and OpenAI have been scrutinized, with Nvidia's CEO defending the company's equity stakes as a minor portion of overall capital raised.

Moreover, the influx of foreign capital into U.S. data centers, primarily from surplus economies like South Korea and Japan, has contributed significantly to the funding landscape. By 2025, approximately $2.5 trillion in foreign investments were supporting U.S. data centers, echoing patterns that preceded the 2008 housing bubble.

Investors are advised to closely monitor these dynamics, particularly the implications of self-referential revenue streams and the potential for distorted demand signals. As the AI sector continues to evolve, the sustainability of these circular funding models warrants careful scrutiny.