FreshSynapse

AI trends & daily updates, in plain English

AI Business

Anthropic Retracts Its Own IPO Pricing Hours Before the Roadshow — the $69.3 Billion Figure Was Wrong

By FreshSynapse Editorial · · 4 min read

A trader in a blue striped shirt pointing at a candlestick chart on one of several monitors at a desk

Photo: Hippopx (public domain). Illustrative image.

What happened. Anthropic’s planned stock-market listing hit an embarrassing snag on Monday 9 November 2026, hours before its investor roadshow was due to begin. According to Reuters, which reviewed the emails, the company circulated a copy of a securities filing to investment banks showing IPO pricing — then sent a follow-up email saying those pricing details were inaccurate and that an updated filing would follow. Anthropic declined to comment to Reuters, and the follow-up email did not say which details were wrong. Reuters said it could not determine what was incorrect.

What the retracted document showed. The circulated copy listed about 866.7 million Class A shares at $70 to $80 each. At the top of that range, the offering would have raised $69.3 billion in gross proceeds — among the largest IPOs on record, though still below the roughly $75 billion SpaceX raised in June. Of the shares, about 843.3 million were to be sold by Anthropic itself and 23.3 million by existing shareholders in a secondary offering. The same document also described a separate $26 billion private placement involving Google, Nvidia, AMD, Broadcom, Micron, Microsoft and Taiwan Semiconductor Manufacturing Company, according to coverage of the circulated copy. Every one of these figures comes from a document Anthropic itself says was wrong, so none of them should be treated as the terms of the offering.

What is actually confirmed. Very little about the terms. Anthropic’s public prospectus — the S-1 filed with the US Securities and Exchange Commission on 26 October, amended on 2 November — leaves the price range and share count blank. What it does confirm is a planned listing of Class A shares on the Nasdaq Global Select Market under the ticker ANT. Reuters reports the roadshow is expected to begin on Tuesday 10 November, with pricing the following week. The Wall Street Journal, as summarised by TechCrunch, reported that Anthropic plans to set a valuation range this week and start trading on Thursday or Friday of next week, around 19 or 20 November.

The valuation slide behind the confusion. The retraction caps a month of shrinking numbers. Anthropic entered October with a floated $2 trillion valuation; by mid-October that had fallen to about $1.5 trillion, and CNBC reported on 9 November that the company is now targeting $1 trillion to $1.5 trillion. The Financial Times reported on 7 November that the company hoped to raise $80 billion to $100 billion at a $1.25 trillion to $1.75 trillion valuation — figures that, like the $69.3 billion, were attributed to investors rather than confirmed by the company. The retracted filing’s $69.3 billion sits below even the bottom of that range, which may be part of why it was pulled — but Anthropic has not said so, and that is speculation.

Why it matters. Anthropic is not just any listing: it would be the biggest AI IPO of the year, and its prospectus discloses about $895 billion in long-term computing commitments through 2045. The company is raising money to pay for the data-centre build-out behind Claude — the same week it announced Claude Fable 5.2 and Mythos 5.2, and a fortnight after reports it had committed about $3.4 billion to gas-fired generators for future data centres. A pricing error circulated to banks on the eve of the roadshow is a bad look for a company asking public investors for tens of billions, and it leaves the market guessing at the real terms until the updated filing lands. Until then, any figure quoted as Anthropic’s IPO size — including $69.3 billion — is unconfirmed.

Sources

  • Anthropic
  • IPO
  • Claude
  • stock market
  • AI business

Related articles