📊 Full opportunity report: The Neocloud Cartel: How the AI Industry Started Renting Compute From Itself on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
In 2026, a small network of AI companies rent compute from each other, creating a cartel centered around Nvidia. This setup gives a few firms control over AI infrastructure but also introduces vulnerabilities.
In 2026, the AI industry has shifted toward a system where the majority of compute power is rented rather than owned, with a small group of companies forming a tightly interconnected cartel centered around Nvidia. This development significantly influences who controls AI infrastructure and how the industry scales.
Recent reports reveal that almost none of the leading AI firms own the GPUs they run on; instead, they lease from a handful of providers, creating a tightly knit network often described as a cartel. CoreWeave, a major GPU rental firm, has a backlog exceeding $55 billion, while companies like Meta and OpenAI have committed tens of billions in leasing agreements.
In May 2026, xAI, a frontier AI lab, became a notable player by leasing its supercomputer to competitors like Anthropic and Google, signaling that even AI labs are now acting as landlords. These leases often include clauses that give the lessor control over capacity, effectively turning hardware into a governance tool.
The core of this system is Nvidia, which supplies the majority of the GPUs and holds equity in many rental firms. Nvidia’s investments, including a $100 billion fund for OpenAI, and its control over chip allocations, give it disproportionate influence over the entire AI compute ecosystem. This concentration of power has led to a circular flow of money, chips, and contracts among a small set of firms, forming a de facto cartel.
The Neocloud Cartel
Almost no one racing to build AI owns the machine it runs on. They rent — increasingly from each other — and the money loops back to one chip maker that’s also an investor in nearly everyone at the table.
The cartel isn’t a conspiracy — it’s the endpoint of extreme capital intensity, real scarcity, and one dominant supplier. But the same circularity that makes it powerful makes it a fuse: each cancelled order is someone else’s missing revenue. Don’t be a price-taker at the bottom of a loop you don’t control — own your inference, keep an open-weight fallback, diversify silicon.
Implications of a Small Group Controlling AI Compute
This emerging cartel consolidates power within a few firms, primarily Nvidia, shaping the future of AI development and deployment. The control over compute resources affects pricing, access, and innovation pace, potentially limiting competition and creating systemic vulnerabilities. The reliance on leasing and contractual control also means that the supply chain is fragile; a disruption or policy change could rapidly impact AI progress.
NVIDIA Tesla L4 24GB PCIe Graphics ACELLERATOR HH/HL 75W GPU 900-2G193-0000-000
24GB Video Memory
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Growth of the ‘Neocloud’ and Its Role in AI Scaling
The ‘neocloud’ category emerged around 2024-2025 as companies faced GPU shortages, prompting a shift toward renting hardware instead of owning. Major players like CoreWeave, Meta, and OpenAI have heavily invested in GPU leasing, with Nvidia dominating supply and financing. This trend has accelerated, culminating in the formation of a tightly interconnected financial and supply network by 2026, with leasing agreements often exceeding hundreds of billions of dollars.
The development reflects a broader industry move: ownership of hardware has become decoupled from AI development, with leasing and contractual arrangements creating a new power dynamic. The emergence of AI labs as landlords marks a fundamental shift in how compute resources are controlled and distributed.
“The cost of a gigawatt of AI data center capacity is roughly $50 billion, with about $35 billion flowing to Nvidia.”
— Jensen Huang, Nvidia CEO
AI compute leasing hardware
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unclear Risks and Potential Disruptions to the Cartel
It remains uncertain how fragile this cartel is in practice. While concentration of power offers advantages, it also presents vulnerabilities; a major supply chain disruption, regulatory intervention, or a shift in chip manufacturing could destabilize the system. The long-term sustainability of this leasing model and its impact on competition and innovation are still emerging topics.
Local AI on Linux in Practice: Build Private LLM Servers, GPU Workstations, Ollama Apps, Dockerized AI Services, and Self-Hosted AI Infrastructure with CUDA, ROCm, vLLM, and Open WebUI
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps for Industry Regulation and Market Stability
Industry analysts expect increased scrutiny from regulators due to the concentrated control over AI compute resources. Companies may also seek to diversify supply chains or develop alternative hardware solutions to reduce dependence on Nvidia. Monitoring how these dynamics evolve will be crucial for understanding future AI development and market stability.
enterprise AI hardware leasing
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why do AI companies prefer renting compute rather than owning hardware?
Due to GPU shortages and high costs, renting provides a flexible, faster way to scale AI models without long-term capital investment. It also allows companies to access the latest hardware without ownership burdens.
How does Nvidia maintain its control over AI compute resources?
Nvidia dominates supply with its chips, holds equity in many rental firms, and controls chip allocation, which effectively grants it leverage over the entire ecosystem.
What are the risks of this compute cartel for the AI industry?
The concentration of power could lead to supply disruptions, limit competition, and create systemic vulnerabilities if any key player faces issues or regulatory intervention occurs.
Could this leasing model hinder innovation in AI?
Potentially, as control over hardware and supply could slow down new entrants or alternative hardware development, impacting overall innovation pace.
What might change in the future to break or reform this cartel?
Regulatory actions, technological breakthroughs, or shifts in supply chain dynamics could challenge Nvidia’s dominance and reshape the compute landscape.
Source: ThorstenMeyerAI.com
