A setback involving Oracle's massive New Mexico AI campus is putting fresh attention on the debt financing behind the world's rapidly expanding data-center buildout.
Oracle's decision to issue a force-majeure notice over potential power delays at a major New Mexico data-center project has renewed concerns about the financing risks surrounding the AI infrastructure boom. Around $18 billion of loans tied to the Oracle-leased project had already come under pressure, highlighting how construction delays, electricity constraints and rising borrowing needs can affect the economics of AI data centers.
The artificial intelligence boom is creating an enormous demand for computing infrastructure. Behind the software and AI models are billions of dollars being spent on chips, electricity, cooling systems, buildings and data-center networks.
Much of that infrastructure is being financed through debt. Oracle has become one of the most closely watched examples because its AI expansion requires enormous capital spending while the company is simultaneously trying to convert future cloud demand into actual cash flow.
What Happened in New Mexico?
On September 24, Reuters reported that Oracle had issued a force-majeure notice to a Blue Owl unit developing a major data-center campus in New Mexico. The notice was linked to potential delays in securing the power required for the project.
The project, known as Project Jupiter, is part of Oracle's broader agreement to provide computing capacity for OpenAI. Reuters reported that the campus is being developed on roughly 1,400 acres in Doña Ana County.
Blue Owl said the notice did not change the parties' commitment to the project, while Oracle has maintained that the project remains part of its AI infrastructure plans.
Why the Debt Matters
The New Mexico project is significant because it demonstrates how AI infrastructure is increasingly being built with large amounts of borrowed money.
Reuters reported on September 18 that approximately $18 billion in loans connected to the Oracle-leased data center had been quoted at roughly 89 to 91 cents on the dollar by syndicate banks including Santander and Jefferies.
That does not mean the project has defaulted or that lenders have lost the money. Debt trading below face value can instead reflect investors demanding compensation for greater uncertainty about the project's timing, economics or risk.
Power Has Become a Financial Risk
AI data centers require extraordinary amounts of electricity. As companies race to build new facilities, obtaining power can become just as important as securing land, chips and financing.
A delay in connecting a data center to the grid can push back the moment when the facility begins generating revenue. But the financing costs associated with the project can continue accumulating during the delay.
This creates a basic financial problem: the infrastructure may require billions of dollars before it can produce the computing capacity that customers have agreed to pay for.
Oracle's Bigger AI Financing Strategy
Oracle announced earlier in 2026 that it expected to raise between $45 billion and $50 billion during the year to expand Oracle Cloud Infrastructure. The company said the financing would combine debt and equity and was intended to support capacity for customers including OpenAI, Meta, NVIDIA, AMD, TikTok and others.
Oracle said the financing plan was designed to maintain an investment-grade balance sheet while funding its rapid cloud expansion.
The challenge is that the amount of capital required for AI infrastructure has grown exceptionally quickly. Investors therefore have to consider not only whether demand for AI computing exists, but also how quickly that demand can translate into revenue and cash flow.
The Risk Is Bigger Than Oracle
Oracle is not the only company borrowing heavily to support AI infrastructure. Major technology companies and specialist data-center operators are also raising large amounts of capital to finance new facilities and computing equipment.
Reuters reported earlier this month that AI-related debt issued during 2026 had reached nearly $500 billion through early August, according to Goldman Sachs analysts. AI-related borrowing accounted for a much larger share of higher-rated U.S. corporate debt issuance than it did just two years earlier.
Investors have increasingly demanded higher compensation for some AI-related debt as concerns grow over construction delays, electricity constraints, supply-chain problems and the sheer volume of new borrowing.
Why Investors Are Watching Closely
The central question is not whether AI will require more data centers. Demand for computing capacity is clearly driving enormous investment.
The bigger financial question is whether the revenue generated by that infrastructure will grow quickly enough to justify the enormous capital expenditure and debt required to build it.
If projects are completed on time and customer demand remains strong, debt-financed infrastructure can generate long-term returns. If projects experience significant delays or demand develops more slowly than expected, however, lenders and investors could face greater risks.
Creative Financing Is Adding Another Layer
The AI infrastructure boom has also produced increasingly sophisticated financing structures. Some projects are financed through special-purpose vehicles, while technology companies can provide guarantees or other forms of support without carrying every liability directly on their own balance sheets.
Such structures can make it easier to raise the capital needed to build infrastructure, but they can also make the overall financial exposure more difficult to understand. The Financial Times reported this month that major technology companies had provided as much as $300 billion in residual-value guarantees connected to AI data centers and chips over the previous year.
These arrangements do not automatically represent a financial crisis. They do, however, show how much financial engineering is now supporting the physical infrastructure behind the AI boom.
This Is Not Yet a Systemic Crisis
Concerns about AI data-center debt should not be confused with evidence of a broad financial collapse. Current developments show growing investor caution and pressure on individual projects, but they do not establish that the banking system or AI industry is facing a systemic debt crisis.
The distinction matters. Large technology companies have substantial businesses, valuable intellectual property and significant access to capital. The risks are concentrated around the scale and speed of infrastructure investment and whether future AI revenues can justify today's spending.
The Next Test for the AI Infrastructure Boom
Oracle's New Mexico project illustrates a problem that could become increasingly important across the industry: AI infrastructure needs more than money. It needs electricity, permits, equipment, construction capacity and customers capable of supporting the resulting contracts.
As borrowing increases, every delay becomes more financially significant. That means investors are likely to pay closer attention to project timelines, customer commitments, power availability, financing structures and the cash flow generated by AI businesses.
Oracle's latest data-center setback does not prove that an AI debt crisis is coming. It does show why the next phase of the AI boom will be judged not only by model breakthroughs, but by whether the enormous infrastructure built to power them can generate enough revenue to support its financing.
Technology, business, finance and the forces shaping the future.
