Anthropic’s IPO Filing Pits Explosive Growth Against an $895 Billion AI Bet
Anthropic’s IPO Filing Pits Explosive Growth Against an $895 Billion AI Bet
Anthropic’s original IPO prospectus, filed last week, offered investors only a partial view of the business. Monday’s amended S-1 added third-quarter figures—and turned the coming listing into a sharper test of whether breathtaking demand can justify an equally breathtaking infrastructure wager.
The new numbers show momentum. Anthropic generated nearly $17 billion in third-quarter revenue, up from $11.6 billion in the prior quarter and more than 12 times the level a year earlier; revenue for the first nine months reached $33.24 billion. But the filing also reported a $23.29 billion quarterly net loss and a $104 billion loss for the nine-month period, much of the latter tied to changes in the fair value of convertible notes and derivative liabilities.
The immediate pressure point was compute. Anthropic swung from $974 million in operating income in the second quarter to a $567 million operating loss in the third, saying the result was “primarily driven by $5.0 billion of compute under short-term commitments.” Compute and infrastructure charges reached $9.9 billion in the quarter.
Then comes the longer-term bet: roughly $895.4 billion in purchase commitments and leases, substantially all for compute and infrastructure. About $447.5 billion is expected by the end of 2030, nearly $70 billion within the following 12 months, with the balance extending through 2045. Anthropic had $131.8 billion in cash and short-term investments at the end of September, but the commitments underline how much future demand it must convert into durable cash flow.
The filing also lays bare the cost of building the company itself. Personnel expenses hit $1.7 billion for 5,300 workers in the quarter—more than $322,000 per employee—with stock-based compensation accounting for $1.1 billion.
Finally, Anthropic paired its financial disclosure with an unusually blunt safety warning. It said future models may “imminently meet” its threshold for dramatically AI-accelerated progress, while insisting no current model has crossed it and that humans must remain meaningfully involved in frontier-model development. Supporters can read that as disciplined transparency; skeptics can reasonably see an IPO-era safety disclosure that also advertises Claude’s ambitions. Either way, the filing makes clear that Anthropic’s growth story and its risk story are now inseparable.
Write a comment