Chip and Cloud Giants Act as Banks to Fund Gigawatt-Scale AI Infrastructure

The race for artificial intelligence capacity has entered a new financial phase where chip and cloud giants are acting as lenders of last resort. To enable the construction of gigawatt-scale data centers, companies like Nvidia and Google are providing credit guarantees to backstop financing that exceeds the risk appetite of traditional commercial banks.

This emerging trend is exemplified by two massive infrastructure plays. In Ohio, Nvidia is in talks to provide a backstop—potentially up to $250 billion, though some reports suggest this figure may be revised downward—to allow OpenAI to borrow against Nvidia's credit for the lease and construction debt of a 10-gigawatt campus in Pike County. The project, developed by SoftBank subsidiary SB Energy, carries a total potential price tag exceeding $500 billion. Similarly, Google is providing financial guarantees for a $15 billion financing plan for an Anthropic data center campus in Hubbard, Texas, developed by Nexus Data Centers.

The Capital Gap

This shift is driven by a fundamental mismatch between the scale of AI ambitions and the constraints of traditional finance. As AI labs move toward 'gigawatt-scale' campuses, the required capital expenditure has dwarfed previous cloud investments. Commercial banks are generally unwilling to shoulder the immense risk associated with such unprecedented infrastructure costs. By stepping in as guarantors, hardware and platform providers are effectively removing the primary bottleneck to physical expansion, ensuring that the infrastructure necessary to run their chips and services is actually built.

A New Power Dynamic

This transition from a simple vendor-customer relationship to a guarantor-tenant partnership fundamentally alters the AI economy. By providing the financial bedrock for these projects, chip and cloud giants are locking in long-term demand and creating deep structural dependencies. While data center operators gain the leverage to build at an unprecedented scale, the giants secure their position as the sole hardware and platform providers for the foreseeable future. This concentration of financial risk and power among a few systemic players accelerates the build-out of AI capacity but ties the physical layer of the internet more tightly to a handful of corporations.

What to Watch

Industry observers are now monitoring whether this 'guarantor model' becomes the standard for all future mega-campuses. The primary remaining question is the final scale of these guarantees; for instance, the exact ceiling of Nvidia's commitment to the Ohio project remains fluid. As these projects move from planning to construction, the market will be watching to see if other hardware providers follow suit or if the financial burden eventually forces a return to more traditional, diversified funding models.