The AI Money Machine: 55% of Investment in AI Companies Comes From Other AI Companies

A 55.2% Figure That Rewrites the AI Funding Story

Most of the money going into artificial intelligence companies comes from other artificial intelligence companies. The Bank for International Settlements examined 1,246 AI firms and found that 55.2% of incoming investments in AI firms came from other AI firms between 2021 and 2025. The finding appears in the central bank group’s bulletin on circular relationships among AI firms. It is a reported research result built from historical deal data, and it carries no forecast or company guidance.

The AI sector supplies the majority of its own outside capital. That changes the question for investors focused on long-term AI spending. The issue becomes how much of the boom is paid for by customers outside the industry, and how much is the industry funding itself.

What the Circular Money Reveals About AI Demand

The investment loop runs in both directions. The same study found that 28.7% of AI firms’ investment deals (by deal value) involved target companies that were also AI firms. AI companies are both the largest pool of buyers for AI equity and active buyers of AI equity themselves.

The financing is also tied to sales. Of all AI-to-AI investment deals in the period, 16.1% by deal count and 46.4% by deal value also involved commercial supply chain relationships. Within those, 64% involved an AI firm that invests in another AI firm and also supplies it with products. The gap between the deal count and deal value figures shows that the largest transactions are the ones most likely to blend an equity stake with a supply contract.

According to the BIS, suppliers finance customers to gain information about their growth path and to steady demand for expensive inputs, a pattern the bulletin ties to chip makers and cloud providers funding model developers. Customers finance suppliers to lock in critical, limited inputs and to keep those suppliers from serving competitors. According to the bulletin, these mechanisms are stronger in AI because of information gaps between firms, highly customized inputs, and a small number of providers [Source: bulletin-137-circular-relationships-among-ai-firms].

Each deal makes sense for the company doing it. Taken together, the deals connect the sector so closely that one company’s revenue can be another company’s investment.

Why Investors Are Already Questioning AI Spending

The finding comes as investors are already considering the cost of the AI buildout. CNBC reported that the market search for balance-sheet cash-cow “quality” stocks is shifting amid mega AI spending. Business Insider reported that investor Michael Burry says he is “more confident than ever” of an AI reckoning in 2027.

Bear Case: Demand That Partly Pays for Itself

The 55.2% figure supports caution. According to the BIS, circular relationships make reported demand partly endogenous to firms’ own financing decisions rather than organic final demand. When a supplier’s capital helps fund a customer’s purchases, some of the resulting revenue links back to the supplier’s own balance sheet, and sales growth can look stronger than the underlying demand from businesses and consumers at the end of the chain.

The second risk is correlation. The bulletin warns that these ties increase correlations between commercial and financial exposures during stress periods. A company that is both an investor in and a supplier to a struggling AI firm takes two losses at once: a markdown on its stake and a drop in orders. With 46.4% of AI-to-AI deal value tied to supply relationships, that double exposure is built into a large share of the sector’s dealmaking.

The third risk is visibility. The BIS points to off-balance-sheet commitments such as residual value guarantees that only materialize during downturns when guarantors are least able to absorb them. These obligations can stay out of view in strong markets and then show up all at once when conditions weaken, leaving retirement savers who depend on earnings quality with little warning from standard disclosures.

Taken together, the data shows an industry where funding, sales, and guarantees run through the same small group of companies. That arrangement works while capital keeps flowing. It leaves less room for error if capital pulls back. Riding a mania is one thing; planning the exit is another, and we wrote a free handbook on both parts here.

Bottom Line for Long-Term Investors

The 55.2% figure gives investors a clear measure of how much the AI boom is financed from within. No scheduled event is on the calendar. The test to watch is whether demand from customers outside the AI sector grows fast enough to replace capital the industry now supplies to itself. Until outside customers carry more of the load, part of the sector’s revenue rests on its own funding.

Contact [email protected] for any questions or corrections.