The AI infrastructure boom has created a new kind of gold rush. Startups, cloud providers, and chipmakers are racing to secure computing power fast enough to keep up with exploding demand for artificial intelligence. In the middle of this scramble stands CoreWeave, a fast-growing AI cloud company, and its CEO, who recently spoke out to defend what critics are calling “AI circular deals.”
While the term may sound technical, the debate behind it cuts straight to how AI companies are financed, how risk is shared, and whether the current AI economy is being built on solid ground—or creative accounting.
In simple terms, AI circular deals happen when multiple companies—often a cloud provider, a chip supplier, and an AI startup—enter agreements that loop money and resources between them. For example, a chipmaker might invest in an AI company, which then uses that funding to buy cloud services powered by the same chipmaker’s hardware. On paper, revenue flows in a neat circle.
Critics worry that these arrangements inflate demand and revenue numbers, making growth appear stronger than it really is. Supporters argue they are just modern versions of strategic partnerships that help young industries move faster.
CoreWeave has found itself right at the center of this debate.
CoreWeave operates as a specialized cloud provider focused almost entirely on AI workloads. Unlike traditional cloud giants, CoreWeave built its business around high-performance GPUs, making it attractive to AI labs that need massive compute resources quickly.
As investment in AI exploded, CoreWeave grew at an incredible pace. That growth, paired with partnerships involving hardware suppliers and AI customers, raised questions about whether some deals were truly independent or part of broader, interconnected agreements.
This is where the idea of “circularity” entered the conversation.
CoreWeave’s CEO has pushed back hard against the criticism. According to him, labeling these partnerships as circular deals oversimplifies how infrastructure-heavy industries actually function.
Building AI data centers requires enormous upfront investment. GPUs, energy infrastructure, and networking don’t come cheap. Strategic partnerships, he argues, are not optional—they’re necessary to share risk, secure supply, and scale fast enough to meet demand.
From CoreWeave’s perspective, revenue is real. The compute is delivered, the GPUs are running, and customers are training real models. Whether partners also have equity stakes or long-term contracts doesn’t change the value being provided.
CoreWeave is far from the only company involved in these kinds of arrangements. As AI demand exploded faster than traditional cloud capacity, unconventional deals became more common.
Chipmakers want guaranteed buyers for expensive hardware. AI startups need compute without waiting years for infrastructure expansion. Cloud providers need capital and supply assurance at the same time.
Circular-looking deals often emerge naturally when these needs collide.
The CEO argues that calling them problematic only makes sense if you ignore the underlying economics of AI infrastructure.
Still, investors and analysts are paying close attention. In a market already fueled by hype, any sign of inflated numbers is taken seriously. Some worry that circular deals could mask real customer demand or create fragile business models that collapse if one player pulls out.
CoreWeave’s leadership insists transparency is the key. Contracts are disclosed, revenues are accounted for, and performance metrics reflect actual usage. In their view, criticism says more about unfamiliarity with AI infrastructure than about misconduct.
This discussion isn’t just about CoreWeave—it’s about defining norms for an industry still writing its own rules. AI infrastructure sits somewhere between software, hardware, and utilities, making it hard to apply old financial playbooks.
If circular deals become widely accepted, they could accelerate AI development by lowering barriers to entry. If they’re restricted or discouraged, consolidation may favor only the largest, cash-rich tech giants.
Either outcome shapes who controls the future of AI.
Even supporters acknowledge that these deals carry risk. Interdependence means trouble can spread quickly if one major partner stumbles. A slowdown in AI funding, changes in regulation, or falling demand could expose weaknesses in tightly linked ecosystems.
CoreWeave’s CEO argues that risk exists in any fast-growing sector. The solution, he says, isn’t avoiding partnerships—but managing them responsibly and being honest with stakeholders.
As AI continues to reshape industries, scrutiny will only increase. Regulators, investors, and the public want reassurance that the infrastructure powering AI is sustainable, ethical, and financially sound.
CoreWeave’s defense of AI circular deals suggests the company is betting on transparency and performance to win that trust. Whether that bet pays off will depend on how well these partnerships hold up under long-term pressure.
For now, one thing is clear: building the backbone of the AI revolution isn’t cheap, simple, or straightforward. And the deals that make it possible may look messy—but that doesn’t automatically make them wrong.
The CEO’s stance highlights a growing tension between innovation speed and financial clarity. In a sector moving as fast as AI, traditional labels may fail to capture modern realities.
AI circular deals may be uncomfortable to analyze, but they could also be the scaffolding holding today’s AI boom together. The real test will come not from headlines, but from whether these systems endure when the market cools.
