Anthropic could begin formal marketing for its initial public offering as soon as the week of 9 November, which would put it on course to start trading before Thanksgiving, Bloomberg reported on Thursday, citing people familiar with the plan.
That is a tight window. The US holiday falls on 26 November, and dealmaking effectively halts in the days around it. A roadshow starting on 9 November leaves roughly a fortnight to price and list.
What is actually fixed, and what is not
Almost nothing is fixed. Anthropic filed a confidential draft registration statement on 1 June. Reported expectations put the float at about $2tn and the raise at as much as $100bn, but those are figures attributed to people familiar with the discussions, not announced terms. No price range has been published.
The timetable has moved before. Investors had expected an October listing; the November date lets the company report third-quarter results before it markets, which is the usual reason an issuer waits.
The company it would list as
Anthropic’s own prospectus is unusually blunt about risk. Reuters reported last week that it warns investors its technology could pose existential risks to humanity, and that the company expects around $518bn of cloud, compute and infrastructure obligations in coming years, most of which it cannot walk away from. Its seven founders want 50.1% of the vote while holding roughly 2% of the shares each.
The contrast across the street
OpenAI has gone the other way. It ruled out a listing this year and is instead asking investors for close to $30bn at a $1.4tn valuation, with Sam Altman saying a public offering now would be ill-advised given the safety debate. Two frontier labs with comparable scale have reached opposite conclusions about whether this is a moment to face a public market.
What to watch
A public S-1. Until the confidential filing is flipped and a price range is attached to it, every figure in circulation is an expectation rather than a term, and the November date is a plan rather than a schedule.
