Crux AI Secures $22 Billion Loan for Google AI Chips

Crux AI, a joint venture between Blackstone and Alphabet, is finalizing a $22 billion loan from ten major banks to purchase Google’s custom artificial intelligence chips. This transaction stands out as one of the largest debt financings ever assembled specifically to fund the hardware required for AI computing infrastructure. The move signals a significant shift in how the industry is capitalizing its rapid expansion, moving away from traditional equity-heavy models toward heavy reliance on borrowed capital.

The lending group includes prominent institutions such as Goldman Sachs, Sumitomo Mitsui Banking Corporation, Barclays, BNP Paribas, and Bank of Nova Scotia. These banks are currently working to bring additional lenders into the deal through syndication. Alongside the primary loan, several banks are providing a separate $1 billion revolving credit facility to add liquidity. This financial structure allows Crux AI to secure the necessary hardware while maintaining cash flow flexibility during its build-out phase.

Collateral Strategy Relies on Chip Value

The financing structure is notable because the $22 billion debt is secured by both the value of the chips and Crux AI’s customer contracts. This effectively turns expensive AI computing hardware and its associated contracted revenue into collateral for one of the largest infrastructure financings of the AI boom. Traditionally, AI infrastructure has been financed through combinations of corporate balance sheets, project finance, and equity investments. However, the extraordinary value of modern AI processors is creating new forms of asset-backed lending where the chips themselves serve as the primary security for the loan.

In this specific case, lenders have an additional source of protection through the venture’s long-term customer commitments. These contracts provide predictable future cash flows, allowing lenders to underwrite the financing more like infrastructure or equipment finance rather than unsecured lending to a technology startup. This approach could become increasingly important as the cost of AI hardware grows into the tens of billions of dollars for individual projects, providing a safer path for banks to enter the market.

Expansion of Google Chip Ecosystem

Crux AI will make significant use of Google’s Tensor Processing Units, or TPUs, which are specialized processors designed to accelerate machine learning and other AI workloads. Unlike traditional cloud infrastructure dominated by general-purpose CPUs and GPUs, this venture relies heavily on Google’s proprietary silicon. This gives Alphabet another avenue for expanding the ecosystem around its custom chips beyond its own internal operations and Google Cloud services, potentially increasing the market share of its specialized hardware.

The venture emerged from a cloud infrastructure initiative announced in May, with Blackstone committing an initial $5 billion of equity. The plan calls for approximately 500 megawatts of data center capacity to come online in 2027, with additional expansion expected afterward. The combination of this initial equity and the new chip loan illustrates how quickly financing requirements are increasing as AI infrastructure moves from individual data centers toward much larger computing platforms.

Trade-offs in Infrastructure Financing

While the loan structure provides immediate capital, it introduces significant financial leverage. The debt could eventually be refinanced with longer-term capital from institutional investors through the investment-grade bond market, giving banks a potential path to syndicate or replace portions of the initial financing. However, this strategy ties the venture’s financial health closely to the sustained demand for its computing capacity and the continued market value of the underlying hardware. If demand for AI computing capacity slows or the value of the chips depreciates faster than expected, the debt burden could strain the project's operational flexibility.

For Blackstone, this deal extends the firm’s already significant exposure to data centers and digital infrastructure. As the world’s largest alternative asset manager, it has increasingly focused on the enormous capital requirements associated with AI, including real estate, electricity, and supporting infrastructure. The scale of this loan demonstrates how these investments are creating new opportunities for banks, but it also highlights the growing complexity of financing the next generation of AI platforms. The shift toward asset-backed lending for chips marks a broader evolution in how the AI buildout is being funded, balancing rapid deployment against long-term financial risk.