Wow, the International Monetary Fund (IMF) has issued a major warning: this circular financing scheme in the AI industry has reached a staggering $1.4 trillion. Large companies and chipmakers invest in downstream AI clients, even guaranteeing their debts. After receiving the funds, clients then use them to purchase chips and cloud services from these large companies. The funds circulate back into the revenue of these large companies, forming a closed-loop structure. Currently, 17 such AI circular transactions have been exposed in the market, and we've compiled them; the scale is staggering: 1. Nvidia ↔ OpenAI: Nvidia invested $30 billion in OpenAI and provided guarantees of up to $105 billion for a data center leasing project in Ohio; OpenAI, in turn, purchased large quantities of Nvidia chips to fill this data center. 2. Oracle ↔ OpenAI: OpenAI and Oracle signed a $300 billion cloud service deal, and Oracle is heavily leveraging its debt to secure this order. 3. Microsoft (MSFT) ↔ OpenAI: Microsoft holds approximately 27% of OpenAI, and OpenAI has committed to investing $250 billion in Microsoft Azure cloud in the future. 4. SoftBank ↔ OpenAI: SoftBank invested $64.6 billion in OpenAI, and OpenAI leases data center resources from SoftBank subsidiary SB Energy. 5. Amazon (AMZN) ↔ OpenAI: Amazon plans to invest up to $50 billion in OpenAI, and OpenAI simultaneously increases its Amazon Cloud partnership to $100 billion. 6. AMD ↔ OpenAI: AMD provided OpenAI with warrants for up to 160 million shares at an exercise price of only $0.01, practically free; in exchange, OpenAI would purchase AMD chips. 7. Google (GOOGL) ↔ Anthropic: Google will invest up to $40 billion in Anthropic, and Anthropic has committed to spending at least $121 billion on Google Cloud. 8. Amazon ↔ Anthropic: Amazon will invest up to $33 billion in Anthropic, and Anthropic will commit to a minimum of $110 billion in spending on AWS. 9. Broadcom AVGO ↔ Anthropic: Banks will lend $63 billion to a special purpose vehicle (SPC) to purchase Broadcom chips and then lease them to Anthropic; Broadcom will also guarantee the residual value of these chips. 10. Microsoft + Nvidia ↔ Anthropic: The two companies will invest a combined maximum of $15 billion, and Anthropic will commit to spending at least $44.9 billion on Microsoft Cloud. 11. SpaceX SPCX ↔ Anthropic: SpaceX will borrow $40 billion to purchase Nvidia graphics cards and then sell the computing power to Anthropic; incidentally, Nvidia is a SpaceX shareholder. 12. Nvidia ↔ CoreWeave CRWV Nvidia holds approximately 11% of CoreWeave and has pledged to buy back $6.3 billion worth of unsold computing power. CoreWeave's chip supply naturally comes from Nvidia. 13. Nvidia ↔ NeoCloud Nvidia has pledged to purchase $36 billion worth of computing power from this mid-sized AI cloud provider. The cloud provider used this guaranteed revenue as collateral to secure loans to purchase Nvidia GPUs. 14. Microsoft ↔ Enscale, Nebulus NBIS, Iren IREN Microsoft signed cooperation agreements with these companies for up to $70.9 billion. These companies will primarily use the funds to purchase GPUs. 15. Google ↔ Former Bitcoin Mining Companies Google provided billions of dollars in guarantees for data center leases by TeraWulf and Cipher Mining CIFR in exchange for equity in the corresponding companies. 16. Meta vs. Blue Owl: A joint venture between the two companies to build a $27.3 billion data center project, with Meta providing a $28 billion guarantee. 17. Anthropic IPO: Nvidia, Google, AMD, Micron, Broadcom, and Microsoft, among other chip and cloud vendors, collectively subscribed to $26 billion worth of Anthropic IPO shares; however, Anthropic itself owes these companies hundreds of billions of dollars. Why is this worth paying close attention to? With the constant circulation of funds and the continuous rolling of debt, semiconductor, cloud vendors, and AI-related companies easily deliver better-than-expected financial reports each quarter, with revenue appearing to surge. But the underlying logic is fragile. This $1.4 trillion funding game ultimately depends on end customers' continued willingness to pay for AI services to sustain it. If downstream AI monetization falls short of expectations, the first to succumb will be AI model companies and emerging cloud service providers that rely on borrowing to expand capital expenditures. Following this, large tech companies that have provided equity investments or lease guarantees to these companies will be dragged into a chain reaction of defaults, which is the core of the IMF's recent warning. When total funds are flowing smoothly, it looks like a perpetual motion machine of profit that never stops; But once cash flow dries up, it could trigger a massive chain reaction of losses. When examining the revenue of AI companies, it is crucial to thoroughly investigate the scale of hidden guarantees and liabilities.