Situation Report
Anthropic is likely to get a high “junk” credit rating after its IPO. That’s the forecast Blackstone gave investors in the $63 billion deal that pays for Anthropic’s chips, according to TechCrunch AI, which cites a Financial Times report. It’s a sharp change. Blackstone used to say the Claude maker would earn an investment-grade rating before 2027.
Now Blackstone expects a rating in the double B range. That’s the top tier of speculative grade, sitting one step below investment grade. Anthropic and Blackstone both declined to comment to TechCrunch.
Threat assessment: moderate. This won’t sink anything, but it makes borrowing more expensive for one of the biggest spenders in AI.
What Changed
- The forecast reversed. Blackstone went from “investment grade before 2027” to “high junk after the IPO.” Blackstone isn’t a bystander here. It anchored the chip financing deal.
- The timing is awkward. Anthropic filed to go public earlier this week. The forecast lands just as public investors start looking at the numbers.
- The spending is the issue. TechCrunch AI notes that Anthropic has signed big compute deals throughout the past 12 months. That spending has already worried investors.
Why the Rating Matters
Credit ratings tell lenders how risky a borrower is. Investment grade (BBB- and above) means low risk, cheap debt and access to the widest pool of buyers. Many pension funds and insurers can only hold investment-grade bonds.
Junk ratings mean higher interest rates and fewer buyers. A double B rating is the best kind of junk, but it’s still junk.
For Anthropic, this matters in a practical way. Frontier AI labs run on borrowed and raised money because training and serving models eats huge amounts of compute. When the cost of capital goes up, every chip, every data center lease and every training run costs more. As TechCrunch AI puts it, a high junk rating “would make it more difficult and costlier” for Anthropic to fund the spending its business is built on.
Context
A junk rating isn’t a death sentence. Tesla and Netflix both spent years in speculative territory while burning cash to grow, and both eventually reached investment grade. Ratings agencies care most about steady cash flow and manageable debt. Fast-growing companies that spend ahead of revenue rarely meet that bar early on.
What stands out is the gap between what Blackstone expected then and what it expects now. A major backer cutting its own outlook shows how the math on AI infrastructure looks to the people funding it. Revenue is growing fast across the frontier labs. Compute commitments are growing faster.
It’s also worth remembering what this is: a prediction from one investor, not an actual rating. The agencies haven’t published anything yet.
Tactical Implications
- For investors watching the IPO: Expect debt costs and compute commitments to get close scrutiny in the filing and on the roadshow. Watch for how Anthropic frames its path to profitability.
- For the wider AI sector: Anthropic’s IPO is a test case. If the first major frontier lab to list gets rated junk, other labs planning to raise money may face the same pricing pressure.
- For enterprise customers: Don’t expect any immediate change. Claude’s pricing and roadmap won’t move on a rating forecast. Over the long run, though, higher capital costs could affect how hard Anthropic pushes on price and capacity.
- For lenders and chip suppliers: Big compute deals backed by a speculative-grade borrower carry more risk. Expect tighter terms or more creative structures, like the Blackstone-anchored chip deal itself.
Outlook
The actual ratings will come after the IPO. Until then, Blackstone’s forecast is the clearest signal yet that Wall Street sees Anthropic’s growth story and its spending as two separate questions. The next few months will show which one public markets care about more. Full details are available in the original TechCrunch AI report.