Bank of America, Blackstone, Citi and Morgan Stanley lend $63 billion to a new company that buys Broadcom chips and leases them to Anthropic. Broadcom guarantees the senior debt, may also lend Anthropic the rent, and already backstops a $35 billion deal from June. Who lends to whom, what secures the money, who carries the risk if AI usage slows, and what it means for companies and users.
Browsing Tag
IPO
4 posts
Anthropic’s S-1, Read Critically: Seven Things the Prospectus Says That the Headlines Missed
A critical reading of Anthropic’s 415-page S-1, every point tied to a page: a founder share that never dilutes and a trust paid by the company it oversees, an $80.7 billion loss that prices deals with Amazon and Google, suppliers that finance their own customer, a compute bill of hundreds of billions with the total left blank, margins that depend on where costs are booked, executive pay tied only to the IPO, and where the filing is more candid than any tech prospectus before it.
Anthropic Files to Go Public. Its Own Prospectus Warns About the Technology It Sells
Anthropic’s S-1, read from the filing itself: $16.3 billion in first-half revenue, a first operating profit in Q2, an $80.7 billion mostly non-cash loss, compute contracts in the hundreds of billions with no total printed, a trust and a 50.1% founder share in control, and risk factors that warn of catastrophic AI risk. Press numbers checked against the prospectus, our May predictions graded, and what it means for companies and users.
Anthropic Puts 5% of Itself Into a Public Benefit Fund. The Real Story Is What It Hasn’t Decided Yet.
Anthropic is committing 5% of its shares, about $48 billion at its last valuation, to a new public benefit fund ahead of its IPO. It chose an LLC over a foundation. Here is what that means, how it compares with OpenAI’s foundation, and the five questions still open.