$22 Billion AI Chip Loan Fuels AI Expansion
Crazyrites Editorial Team
September 17, 2026
A group of 10 banks is providing a reported $22 billion AI chip loan to support Crux AI, a new cloud computing venture created by Blackstone and Alphabet as demand for artificial intelligence infrastructure continues to grow.
The financing is one of the latest examples of how banks and investors are becoming increasingly involved in funding the expensive hardware needed to run advanced AI systems.
According to Reuters, the $22 billion AI chip loan will be used to purchase Google’s custom Tensor Processing Units, or TPUs. The debt will reportedly be backed by the value of those chips and Crux AI’s customer contracts.
The development highlights a major change in the AI industry. The race to build better AI models is no longer only about software and research. Companies also need enormous amounts of computing power, data centers, electricity and specialized processors.
What Is the $22 Billion AI Chip Loan?
The $22 billion AI chip loan is a financing arrangement being provided by a group of 10 banks for Crux AI.
The lenders reportedly include Goldman Sachs, Sumitomo Mitsui Banking Corp, Barclays, BNP Paribas and Bank of Nova Scotia. The banking group is also working to bring additional lenders into the deal through syndication, according to reports.
The loan is particularly notable because it is designed around AI processors rather than simply financing the construction of traditional infrastructure.
Google’s TPUs are specialized processors developed for AI workloads. They are designed to handle tasks such as training and running advanced AI models.
The financing therefore connects the value of the physical chips with the expected revenue generated from customers using Crux AI’s computing services.
| Key Detail | Information |
|---|---|
| Financing | $22 billion |
| Venture | Crux AI |
| Main partners | Blackstone and Alphabet |
| Main hardware | Google Tensor Processing Units |
| Banks involved | 10 banks |
| Blackstone equity commitment | $5 billion |
| Initial planned capacity | 500 megawatts |
| Target for initial capacity | 2027 |
| Separate credit facility | Reported $1 billion |
| Main purpose | AI computing infrastructure |
What Is Crux AI?
Crux AI is a new cloud venture created through a partnership between Blackstone and Google.
Blackstone announced the partnership in May 2026, saying it would make an initial $5 billion equity commitment to help bring 500 megawatts of data center capacity online in 2027.
Google is contributing its TPUs, software and services to the venture.
Blackstone said the new company would provide data center capacity, operations, networking and Google Cloud TPUs as a compute-as-a-service offering.
The $22 billion AI chip loan now adds a major debt component to that business model.
Crux AI is intended to serve customers that need large amounts of computing capacity, including AI laboratories, technology companies, enterprises and governments.
Instead of every customer having to build its own large AI infrastructure, the company can provide access to computing resources through the cloud.
Why AI Chips Need So Much Money
The $22 billion AI chip loan shows how expensive modern AI infrastructure has become.
Advanced AI systems require large numbers of powerful processors.
These processors must operate inside data centers that also require cooling systems, networking equipment, electricity and physical space.
As AI models become larger and more widely used, demand for computing capacity has increased.
This has created a new investment cycle involving technology companies, infrastructure investors, banks and other financial institutions.
The cost is not limited to buying chips. Companies also need to maintain the equipment and keep enough electricity available to operate large computing facilities.
As a result, AI infrastructure has increasingly become a major capital investment area.
Why Google TPUs Matter
The $22 billion AI chip loan is closely connected to Google’s Tensor Processing Units.
TPUs are Google’s custom AI processors. Google says its TPUs have been developed and deployed for more than a decade and are used for AI training and inference workloads. They also support Google’s Gemini services and other AI products.
Unlike general-purpose processors, specialized AI chips are designed for particular computing workloads.
This can make them useful for organizations running large AI models.
For Crux AI, access to Google’s TPUs provides an important part of the computing infrastructure that the company plans to offer to customers.
The financing allows the venture to acquire those processors without relying entirely on upfront equity funding.
How the Financing Works
One important feature of the $22 billion AI chip loan is the reported collateral structure.
According to Bloomberg reporting cited by Reuters, the debt will be backed by the value of Google’s TPUs and Crux AI’s customer contracts.
In simple terms, the lenders are not relying only on the overall idea of AI growth.
The financing is connected to specific assets and expected customer payments.
That structure can make it easier for lenders to assess the transaction because there are identifiable assets and contracts supporting the debt.
The banking group is also reportedly considering syndication, which would allow additional lenders to participate and spread the financial exposure across more institutions.
A Separate $1 Billion Credit Facility
The $22 billion AI chip loan is not the only financing linked to the venture.
Reports also indicate that some of the participating banks are providing a separate $1 billion revolving credit facility.
A revolving credit facility gives a company access to funds that can be drawn and repaid according to the terms of the agreement.
This type of financing can provide additional flexibility for a growing business.
For a new AI infrastructure company, having access to additional working capital can be useful as operations expand.
The precise terms of the facility were not publicly disclosed in the reports.
Why Banks Are Funding AI Infrastructure
The $22 billion AI chip loan comes as financial institutions are increasingly looking at AI infrastructure as a large financing opportunity.
AI companies need enormous amounts of capital to expand computing capacity.
Until recently, much of the discussion around AI investment focused on venture capital and technology-company spending.
Now, debt markets and large banks are becoming increasingly important.
Banks can finance equipment purchases and infrastructure while companies preserve some of their equity capital for other uses.
This creates a different way of funding the AI expansion.
At the same time, lenders must consider the risks involved because AI technology is developing quickly and the long-term demand for particular chips or computing services can change.
The Growing Cost of Data Centers
The $22 billion AI chip loan also reflects the rapidly growing cost of building AI data centers.
A modern AI data center requires more than servers.
It needs high-capacity electrical connections, cooling equipment, networking systems, buildings and specialized processors.
Power availability has become particularly important.
Large AI facilities can consume significant amounts of electricity, meaning companies need reliable energy supplies as well as suitable locations.
Crux AI’s planned 500 megawatts of initial capacity illustrates the scale of the infrastructure involved.
Blackstone said in its May announcement that the venture plans to bring this capacity online in 2027 and scale significantly over time.
What This Means for the AI Cloud Market
The $22 billion AI chip loan could help Crux AI compete in an increasingly crowded cloud computing market.
Large technology companies are investing heavily in AI computing infrastructure.
Cloud customers, meanwhile, are looking for access to advanced processors without necessarily purchasing and operating those systems themselves.
This creates an opportunity for specialized cloud providers.
Crux AI’s model is designed to provide dedicated computing infrastructure to customers that require significant AI capacity.
The company says its target customers include AI labs, technology companies, enterprises and governments.
If demand continues to increase, the business could potentially expand its infrastructure beyond the initial 500-megawatt target.
The Role of Blackstone
The $22 billion AI chip loan also shows Blackstone’s growing involvement in technology infrastructure.
Blackstone is one of the world’s largest alternative asset managers and has invested heavily across infrastructure and technology-related areas.
For Crux AI, Blackstone’s initial $5 billion equity commitment provides a substantial capital base.
The additional bank financing creates a much larger pool of capital for purchasing AI processors and supporting the venture’s expansion.
This combination of equity and debt is common in large infrastructure projects because it allows companies to finance major assets without relying entirely on shareholder capital.
The Role of Alphabet
Alphabet’s role in the $22 billion AI chip loan structure is different from Blackstone’s.
Google is providing its specialized TPUs, software and services.
That gives Crux AI access to technology developed within Google’s broader AI infrastructure ecosystem.
Google has spent years developing its own AI chips because large-scale AI workloads require specialized hardware.
The Crux AI partnership gives Google another channel for making its TPU technology available to customers outside the traditional Google Cloud structure.
Blackstone’s announcement said the venture would provide another option for customers seeking access to Google’s cloud TPUs.
What Could Go Right
The $22 billion AI chip loan could help accelerate the development of new AI computing capacity.
If demand for AI services continues to rise, Crux AI could benefit from having large amounts of specialized hardware available.
Customers could gain access to computing resources without having to build their own facilities.
The financing could also allow the company to move more quickly than relying solely on equity investment.
For banks, the transaction provides exposure to a growing infrastructure segment.
For Blackstone and Alphabet, it supports the expansion of a new cloud business built around AI computing.
What Are the Risks?
The $22 billion AI chip loan also carries risks.
AI infrastructure requires huge upfront investment, while technology can change quickly.
A new generation of processors could eventually reduce the demand for older hardware.
There is also the question of whether AI customers will generate enough revenue to justify the enormous cost of computing infrastructure.
Electricity prices and data center construction costs could also affect profitability.
Furthermore, if demand for AI computing slows, companies operating large data centers could face pressure to keep their facilities fully utilized.
The financing structure is intended to provide lenders with protection through chips and customer contracts, but no large infrastructure investment is completely free of risk.
A New Era of AI Financing
The $22 billion AI chip loan is part of a broader shift in how the AI industry is being financed.
AI companies are no longer raising money only to hire researchers or develop software.
They are also financing physical infrastructure on a scale similar to other major capital-intensive industries.
This includes chips, servers, data centers, electricity and networking equipment.
That shift is creating opportunities for banks and institutional investors.
It is also creating new questions about how much debt the AI industry can responsibly support.
As more companies borrow billions of dollars for computing infrastructure, investors will increasingly focus on customer demand, cash flow and the useful life of the hardware.
What Comes Next for Crux AI?
The $22 billion AI chip loan gives Crux AI a substantial financial base as it works toward its first major infrastructure target.
The venture plans to bring 500 megawatts of capacity online in 2027.
The company could then expand its capacity as demand develops.
For now, the focus will be on building the infrastructure, acquiring the required processors and attracting customers.
The success of the project will ultimately depend on whether demand for AI computing remains strong enough to support the cost of the infrastructure.
Conclusion
The $22 billion AI chip loan marks an important development in the financing of artificial intelligence infrastructure.
Blackstone and Alphabet’s Crux AI is using the financing to acquire Google’s specialized TPUs and build computing capacity for AI laboratories, companies, enterprises and government customers.
The deal also demonstrates how the AI boom is creating demand for new forms of financing. Banks are increasingly helping fund the physical technology needed to support advanced AI systems, while infrastructure investors are providing substantial equity commitments.
However, the scale of the financing also highlights the importance of careful investment planning. AI technology is developing quickly, and companies must balance the need for more computing power with the costs of chips, data centers, electricity and debt.
Crux AI’s planned 500-megawatt capacity in 2027 will be an important milestone. If demand for AI computing continues to grow, the venture could become another significant part of the expanding cloud infrastructure market.
For now, the $22 billion AI chip loan provides a clear example of how the next stage of the AI race is increasingly being built not only with software and research, but also with massive investments in physical computing infrastructure.
Sources
- Reuters, “Banks provide $22 billion chip loan to Blackstone, Alphabet AI cloud venture, source says,” September 16, 2026.
- Blackstone, “Blackstone Announces Joint Venture with Google to Create New TPU Cloud,” May 18, 2026.
- CNA / Reuters, reporting on the $22 billion financing and Crux AI’s planned 500 MW capacity.
