Test report DSG-2140 · Rev E · tested October 17, 2026

AI Datacenter InfrastructureDevice under test

Anthropic Offers Blackstone Over 10% Yield in $60B AI Chip Financing

FT reports Anthropic offered Blackstone over 10% yield on $60 billion in AI chip financing, one of the priciest large AI infrastructure debt deals yet. Terms remain undisclosed.

Read
2 min
Words
469
Node
20nm
Operator
Priya Raman

Spec summary

  1. Anthropic offered Blackstone a yield above 10 percent on $60 billion in AI chip financing, per the Financial Times.
  2. The financing would fund AI chips, the accelerators used for machine-learning workloads.
  3. Blackstone would act as lender through private credit channels.
  4. The FT report does not confirm the deal has closed or disclose duration, covenants or security terms.
Anthropic offers Blackstone over 10% yield in $60B AI chip financing - FT reports - Investing.com
Fig. AAnthropic offers Blackstone over 10% yield in $60B AI chip financing - FT reports - Investing.com — Fotorech / pixabay

Anthropic has offered Blackstone a yield above 10 percent on a $60 billion financing package for AI chips, the Financial Times reports. The figure marks one of the highest pricing levels attached to any large-scale AI infrastructure deal to date.

The report offers little beyond the headline terms. That absence is itself informative.

What does the deal cover?

According to the FT, the $60 billion facility would finance AI chips — the specialized accelerators (processors designed for machine-learning workloads) that dominate datacenter capital spending. Blackstone, the private credit manager, would provide the financing. The reported yield exceeds 10 percent.

Key elements of the reported structure:

  • Borrower: Anthropic, the AI developer behind the Claude model family.
  • Lender: Blackstone, acting through private credit channels.
  • Amount: $60 billion.
  • Pricing: a yield above 10 percent.
  • Purpose: financing AI chips.

Why does the yield matter?

A double-digit coupon is expensive money. Companies with strong equity narratives typically borrow far below that threshold in syndicated markets. Anthropic appears willing to pay it.

The pricing tells us two things, and conceals a third. First, lenders price AI compute collateral conservatively; chips depreciate fast and resale markets remain unproven at this scale. Second, Anthropic's cash needs are large enough that it will pay a premium rather than raise equity at an unwanted valuation.

The concealed item is term structure. The FT report does not specify duration, security packages, covenants, or whether the yield includes an equity kicker (additional return paid in shares rather than cash). Any of those could change what the 10 percent figure actually costs Anthropic.

Who else has done this?

AI developers have increasingly turned to private credit for compute buildouts, and asset managers holding trillions in dry powder have moved to fill the gap. Blackstone has been among the most aggressive. A $60 billion commitment to a single borrower would rank among the largest private credit financings on record.

For context, that sum approaches the gross domestic product of a mid-sized European economy. It is roughly the market value of many household-name industrial conglomerates.

What is not said?

The report leaves several questions open. There is no confirmation that the deal has closed. There is no detail on which chip suppliers stand to benefit. There is no disclosure of how the debt sits against Anthropic's existing capital structure, nor of investor demand at the quoted yield.

Until those terms surface, the 10 percent figure should be read as an offer, not a market clearing price. One long-sentence caveat worth holding onto: if a company widely viewed as a frontier AI leader must pay double-digit rates to fund its core hardware, the financing environment for second-tier AI infrastructure players is likely materially harsher.

We will update this story as terms are confirmed.

By Priya Raman

via Google News: AI chip (Source)

Filed under

  • anthropic
  • blackstone
  • private-credit
  • ai-chips
  • ai-infrastructure
Share this article:

More from Priya Raman

Priya Raman

Show full bio

Correspondent covering money, funding and market shifts around hardware layer of artificial intelligence at Die Signal.

407 articles

Same lot · LOT-C1CD

« Previous article