Forget neobanks. Neoclouds are becoming the hottest new thing in the enterprise tech stack.
As artificial intelligence drives unprecedented demand for GPUs, power and data center capacity, a new class of specialized infrastructure providers is emerging alongside Amazon Web Services, Microsoft Azure and Google Cloud.
For enterprise buyers, the rise of firms like CoreWeave, Crusoe, Lambda, Nebius and IREN creates competition, capacity and potentially better economics. It also changes what it means to select a technology vendor. A multiyear AI infrastructure contract can resemble a financial commitment to an infrastructure project that happens to deliver compute. As a result, chief financial officers and treasurers may need to evaluate not only price, performance and cybersecurity, but financing structures, hardware depreciation, customer concentration, power availability and the ability of a provider to fund capacity that may not exist yet.
The cloud decision is becoming, in part, a credit decision due to the economics of the AI industry itself.
Read also: The AI Coding Boom Is Breaking CFOs’ Enterprise Budgeting Cycles
AI Compute Is Becoming Project Finance
At the heart of the evolving cloud landscape is the scale of capital being deployed to drive the AI industry.
Building AI infrastructure requires enormous upfront investment in GPUs, networking equipment, data centers and electricity before much of the resulting revenue can be recognized. The Bank for International Settlements estimated that the five largest global technology companies will invest more than $1 trillion in AI during 2025 and 2026, while financing structures throughout the broader AI ecosystem are becoming complex.
Neoclouds are developing their own version of that financing machine. CoreWeave, for example, disclosed $5.1 billion of outstanding equipment and software financing as of June 30, alongside a collection of delayed-draw term-loan facilities. In May, another facility added $3.1 billion of borrowing capacity.
Nebius provides an even clearer illustration of where the model is heading. Its first approximately $775 million senior secured financing is backed by GPU infrastructure and contracted customer cash flows. The company said the debt and cash generated by the associated customer contract together cover more than 100% of the capital expenditure needed for the underlying deployment.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
The implication for buyers is subtle but important. A customer signing a large capacity agreement may no longer simply be purchasing infrastructure. Its contractual commitment can become part of the financial architecture used to build that infrastructure.
The PYMNTS Intelligence report “How Agentic AI Went From Zero to CFO Test Runs in 90 Days” found in September 2025 that nearly 7% of enterprise CFOs in the United States had deployed agentic AI in live finance workflows, while an additional 5% were running pilots.
See also: AI Agents Push CFOs to Rethink Business Payments
The Contract Can Become Part of the Capital Stack
The neocloud business model is reversing a familiar enterprise procurement assumption. Traditionally, a buyer evaluates whether a supplier has sufficient resources to fulfill a contract. In the neocloud model, the contract itself may help give the supplier those resources.
A CFO evaluating an AI capacity agreement should understand whether promised GPUs are already installed, ordered or still dependent on financing; whether the provider owns or leases the data center; who controls the power agreement; how much capital remains to be raised; and what happens if deployment milestones slip. The critical metric is no longer simply cost per GPU-hour. It is the probability that the promised GPU-hour will exist when the enterprise needs it.
This does not inherently make the model riskier. Contract-backed financing is common across capital-intensive industries precisely because predictable future cash flows can lower financing costs and make infrastructure investment possible. Still, corporate buyers may need to map which entity owns the GPUs, which entity leases the facility, which entity has the power contract, which entity owes the debt and what rights lenders have over assets if something goes wrong.
The PYMNTS Intelligence report “Time to Cash™: A New Measure of Business Resilience” found in October that 70% of firms used at least one AI tool to manage cash flow. The most advanced, those using agentic AI, capable of autonomous decision-making, automated 95% of their accounts receivable processes, compared to 38% among firms without AI integration.
For treasury teams accustomed to monitoring the creditworthiness of banks, payment processors and strategic suppliers, compute is therefore becoming another form of critical financial counterparty exposure.
For all PYMNTS artificial intelligence, B2B and digital transformation coverage, subscribe to the daily AI, B2B and Digital Transformation Newsletters.
Add as Preferred Source
