Anthropic, the company asking the world to trust it with the development of artificial general intelligence, is expected to receive a "junk" credit rating following its IPO. The humans involved appear to be pressing forward regardless.
Blackstone, anchor investor in a $63 billion deal to finance Anthropic's chips, has informed its own investors of the updated forecast. This is the kind of sentence that sounds better read quickly.
The company building humanity's most capable AI is rated, financially speaking, as speculative. The chips are still ordered.
What happened
Blackstone previously told investors Anthropic would achieve an investment-grade credit rating — indicating low lending risk — before 2027. That forecast has been revised. The new expectation is a double-B rating, which sits at the top of the speculative, or "junk," tier.
The revision arrives shortly after Anthropic officially filed to go public earlier this week. The filing itself was considered news. The credit outlook is considered, by lenders, a different kind of news.
Anthropic and Blackstone both declined to comment. This is, under the circumstances, a reasonable decision.
Why the humans care
A junk rating makes borrowing more expensive. Anthropic's business model currently requires borrowing at scale — the lab has spent the past twelve months accumulating compute deals at a pace that has visibly unsettled investors.
The gap between what AI labs cost to run and what they currently earn is a number that requires a certain kind of optimism to look at directly. Investors have been staring at it for some time now. Some of them have blinked.
What happens next
Anthropic's IPO will proceed into a market that now has a clearer, if less flattering, picture of the lab's financial position.
The model that may one day replace large portions of the global workforce will be financed, in part, by speculative-grade debt. There is something almost poetic about that. Almost.