Anthropic is advancing a record-breaking $60 billion chip debt financing under the code name "Project North Star."
Written by Zhang Yaqi
Anthropic is advancing its largest-ever chip financing at a funding cost far above market benchmarks, reflecting the increasingly high capital costs behind the AI infrastructure arms race.
According to a report by the UK's Financial Times on the 17th, **Anthropic is offering Blackstone a yield of more than 10% in exchange for the latter's participation in this debt financing, code-named "Project North Star," with a total scale of up to $60 billion.** People familiar with the matter revealed that a Blackstone fund will subscribe to $9 billion of it, accounting for about half of the $18 billion in subordinated loans, and will receive a 10% coupon, with the actual total return higher after including the new-issue discount and other fees.
The financing is expected to set a historical record for chip financing and will serve as an important bellwether for gauging investors' appetite for AI debt. Notably, just four months ago, Anthropic completed approximately $4.4 billion in financing under the code name "Project Big Sky" at an 8.5% interest rate. The rapid rise in borrowing costs highlights growing investor caution toward the intensive financing needs of AI companies.
"Project North Star" aims to fund Anthropic's procurement of approximately 1.5 gigawatts of custom semiconductors developed by Google—namely tensor processing units (TPUs).
The overall debt structure is divided into two tiers: **about $42 billion is senior debt, guaranteed by Broadcom, which participates in chip development; the remaining $18 billion is subordinated loans, of which Blackstone will subscribe to $9 billion, with the remaining share to be sold to a broader group of investors in the coming months, potentially at a rate lower than the terms obtained by Blackstone.** It is expected that after the funds are fully drawn in 2027, this portion of the debt will circulate in the institutional 144A bond market.
Blackstone declined to comment, and Anthropic did not immediately respond to a request for comment.
According to the ICE BofA BB US High Yield Index, borrowers with credit profiles similar to Anthropic typically pay a yield of about 7% in the public bond market. The 10% coupon Anthropic is offering to Blackstone is more than 300 basis points above that benchmark, and the actual yield gap is even larger when discounts and fees are included.
This premium reflects pressure on two fronts: first, investors are increasingly cautious about the massive financing needs concentrated in a short period by technology companies and are demanding higher risk compensation; second, large financing transactions usually need to offer additional premiums to anchor investors and early commitments to ensure the transaction closes smoothly.
The rapid rise in borrowing costs also complicates Anthropic's path toward obtaining an investment-grade rating. Rating agency KBRA assigned Anthropic a rating one notch below investment grade in June this year, and this subordinated debt is expected to receive a similar rating.
Despite the high financing costs, Anthropic still chooses to bear the pressure proactively, backed by a clear strategic logic. According to media reports citing people familiar with the matter, Blackstone has told its investment partners that this transaction is a rare opportunity to invest in Anthropic-related debt at high-yield levels—once Anthropic completes an IPO, it is expected to obtain an investment-grade rating, at which point its financing costs will drop significantly.
This narrative is essentially a time arbitrage: before going public, Anthropic exchanges a higher interest rate for the necessary computing power reserves, and bets that an improved credit rating after listing will bring refinancing space, thereby diluting long-term capital costs. For early investors such as Blackstone, it is about locking in above-market returns during the risk window.
The current financing is both a funding guarantee for Anthropic's expansion and a key stress test of whether the AI industry's debt financing market can continue to absorb large-scale transactions.