Anthropic's IPO prospectus: what it means for you

Anthropic, the company behind the Claude chatbot, is preparing to sell shares to the public for the first time. A document meant to convince investors to buy in has leaked, and it reveals both fast growth and blunt warnings about the technology Anthropic sells.

Reuters reported on 28 September 2026 that it had obtained Anthropic's IPO (initial public offering, when a private company first sells shares on a stock market) prospectus. Anthropic had confidentially submitted a draft version to US regulators on 1 June 2026 and declined to comment on the leak.

The numbers are striking. Anthropic's 2025 revenue grew twelvefold to almost $4.6 billion, and its revenue run rate — sales projected over a full year based on recent performance — passed $65 billion by the end of July 2026. But the company also posted a $42 billion net loss in 2025, on an operating loss of about $8 billion, and spent $7.33 billion on computing and infrastructure that year. It plans to spend $518 billion more on cloud, computing and infrastructure in the years ahead.

Anthropic is reportedly targeting a valuation above $2 trillion, more than double the roughly $965 billion it was valued at in May, based on projected 2028 revenue of $190–200 billion. The offering could raise up to $100 billion, which would exceed SpaceX's June 2026 listing. It's expected to list on Nasdaq, with Morgan Stanley, Goldman Sachs and JPMorgan running the sale; Nvidia is reportedly considering putting in around $10 billion. Nearly a quarter of Anthropic's revenue comes from just two unnamed customers, and the filing warns that many of its biggest clients aren't tied to long-term contracts.

Buried in the document's roughly 80 pages of risk factors are warnings that AI could pose a "catastrophic or existential risk to humanity," that systems improving themselves could trigger an "intelligence explosion," and that AI could show "self-preserving behaviors" such as trying to "resist shutdown," according to the prospectus reported by Gizmodo.

Michael Field of Morningstar called the growth "truly remarkable," while Kathleen Brooks of XTB said the heavy spending signals more business ahead for chip and infrastructure companies, telling clients "the capex cycle is not over yet," according to Reuters' report via Investing.com.

If you use Claude, this filing doesn't change what the product does today. But it shows how much money Anthropic is burning to build and run it, and how concentrated its paying customers still are. If you're considering buying shares once the IPO happens, the prospectus itself flags the risks bluntly, both financial (customer concentration, heavy losses) and about AI safety more broadly.

The exact listing date isn't confirmed. Some reports point to mid-October, others to November 2026. Anthropic's rival OpenAI is now expected to hold its own IPO in 2027.