Blackstone is quietly assembling another massive debt package, this time worth roughly $36 billion, to bankroll Anthropic's growing appetite for Google's custom chips. The move signals that the private credit boom in AI infrastructure shows no signs of slowing down.
Bloomberg reported the news on August 4. The deal remains in early stages, with Blackstone currently gauging investor interest. If it closes, it would be the second major financing round in less than a year. The first package, a $35-36 billion joint effort with Apollo Global Management, wrapped up between late May and early June 2026.
Both deals follow the same playbook. A special purpose vehicle, or SPV, purchases Google's Tensor Processing Units and leases them directly to Anthropic. This structure keeps the chips legally separate from Anthropic's balance sheet. If Anthropic ever hit financial trouble, investors still own the hardware. It's a financing trick borrowed from real estate and aircraft leasing, now repurposed for silicon.
Why the infrastructure model matters
The SPV approach lets Anthropic expand its data center footprint across multiple US states without the crushing burden of buying billions in equipment outright. The company gets the computing power. Investors get predictable lease payments backed by tangible assets. Everyone wins on paper.
The first deal already generated secondary market activity. A chunk of that $35-36 billion hit the secondary market by July 2026, meaning institutional investors were eager to buy pieces of the deal from initial holders. That appetite suggests a second round makes sense to Blackstone's dealmakers.
The bigger picture for private credit
Combined, both packages would touch $72 billion. Put that against the roughly $1.4 trillion in total annual US leveraged lending, and you start seeing just how much capital is flooding into AI infrastructure right now. Private credit firms are essentially becoming the primary lenders for the compute arms race that powers modern AI.
The secondary market's receptiveness also matters. When investors quickly resell debt into secondary markets, it signals confidence that the underlying deal is solid. No one wants to hold a ticking time bomb.
Final terms, investor commitments, and exact timing remain fluid. But the direction is clear: more capital, more chips, more infrastructure. The race for AI dominance increasingly runs through balance sheets and debt markets, not just R&D labs.
This material is for informational purposes only and should not be construed as financial advice or a recommendation to invest in any particular security or strategy.
