A $22 Billion Bet on AI Chips: Banks Back Blackstone and Google’s New Cloud Venture

The AI Boom Just Borrowed $22 Billion

The race to build artificial intelligence infrastructure has reached another extraordinary milestone. A group of 10 major banks is reportedly providing $22 billion in debt financing to support Crux AI, a new cloud-computing venture created by Blackstone and Alphabet, Google's parent company. Bloomberg first reported the financing, citing people familiar with the matter, and Reuters subsequently reported the details. The money is expected to be used to purchase Google's specialized Tensor Processing Units, or TPUs — the chips designed specifically for demanding AI workloads.

The scale of the financing tells an important story about the AI industry. The biggest challenge for AI companies is no longer simply building clever software. They need enormous amounts of computing power to train and run increasingly sophisticated models, and that means buying chips, constructing data centers, securing electricity and building networks capable of moving huge amounts of data. In other words, the AI race is becoming an infrastructure race — and Wall Street is now putting tens of billions of dollars behind it.

Meet Crux AI

Crux AI was created through a partnership between Blackstone and Google specifically to expand access to AI computing infrastructure. Blackstone announced in May that it would make an initial $5 billion equity commitment, with the venture targeting its first 500 megawatts of data-center capacity by 2027 and plans to expand significantly afterward. Google, meanwhile, will provide its TPUs along with software and services.

The idea is relatively straightforward: instead of every AI company having to build its own enormous computing infrastructure, Crux AI can provide dedicated capacity as a service. Its potential customers include AI laboratories, technology companies, enterprises and governments that need large amounts of accelerated computing power. That puts Crux in the middle of a market that is expanding rapidly as organizations race to develop and deploy AI systems.

Why Are Banks Willing to Lend $22 Billion?

Perhaps the most fascinating part of the deal is not simply the size of the loan, but what is being used to secure it.

According to Bloomberg's report, the debt will be used to purchase Google's TPUs and will be backed by the value of those chips as well as Crux AI's customer contracts. The banking group reportedly includes Goldman Sachs, Sumitomo Mitsui Banking Corp., Barclays, BNP Paribas and Bank of Nova Scotia, with additional lenders potentially joining the financing through syndication.

That structure shows how valuable AI computing hardware has become. A few years ago, financing billions of dollars of specialized processors would have sounded like an unusual proposition. Today, advanced AI chips are among the most strategically important pieces of technology in the world, because without enough computing power, even the most advanced AI models cannot operate at massive scale.

The debt could eventually be replaced by longer-term financing from institutional investors in the investment-grade bond market, according to the Bloomberg report. Some banks are also reportedly providing a separate $1 billion revolving credit facility.

Google Has Its Own Answer to Nvidia

There is another interesting layer to this story: Google's TPUs are becoming an increasingly important alternative to Nvidia's AI processors.

Nvidia has dominated the market for the high-performance GPUs used to train and operate AI models, but Google has been developing its own AI accelerators for more than a decade. Its TPUs power Google's Gemini systems and other large-scale AI workloads. Through Crux AI, Google gets another route to put those processors in the hands of customers outside the traditional Google Cloud model.

That could make the competition for AI infrastructure even more interesting. The industry is no longer simply about which company has the best AI model. It is also about who controls the chips, data centers, electricity and cloud platforms underneath those models.

AI Is Becoming a Massive Infrastructure Business

The Crux AI deal arrives during a period of extraordinary spending on AI infrastructure. Technology companies are committing huge sums to data centers and specialized chips because demand for computing power continues to grow. The financial world is increasingly becoming part of that expansion, providing debt alongside the enormous equity investments flowing into AI companies.

There is, however, another side to the story. The more AI infrastructure is financed with debt, the more important future demand and revenue become. If AI usage grows as aggressively as companies expect, the infrastructure could generate substantial long-term business. If demand fails to meet those expectations, highly leveraged projects could face more pressure. The Bank for International Settlements has already warned that the AI boom is creating new financial-stability risks, particularly as large investments increasingly involve debt financing.

The Real AI Race May Be Happening Behind the Screen

Most people experience the AI revolution through a chatbot, an image generator or an AI-powered app. They see the software. They rarely see the enormous physical infrastructure underneath it.

But deals like this reveal what is happening behind the screen. Every AI response requires computing power. Every new model needs infrastructure. Every increase in usage means more chips, more servers and more electricity. And companies are now building that infrastructure on a scale that requires the financial system itself to participate.

That's why the $22 billion Crux AI financing matters beyond Blackstone and Google. It is another sign that artificial intelligence is becoming one of the largest infrastructure investment stories of the decade. The companies building the future of AI may be writing the software, but increasingly, banks, chipmakers, cloud providers, energy companies and data-center developers are building the physical world that makes that future possible.

And with $22 billion now reportedly lined up for one new AI cloud venture, that physical race is getting very, very expensive.