Surprising things we have learnt from Anthropic's blockbuster share listing

Anthropic has published the paperwork for its long-awaited debut on the US stock market, a 335-page filing that reveals for the first time how the maker of the Claude chatbot earns and spends its money, who controls it, and the risks it sees in its own technology.

First, the risks. Anthropic warns that sufficiently advanced AI systems could, without appropriate safeguards, oversight and security, pose catastrophic or existential risks to humanity. It also says a hacking group it believes was sponsored by China manipulated Claude less than a year ago to intrude on large technology companies, financial institutions, chemical manufacturers and government agencies. Its future, it says, depends on access to the chips and data centres needed to build and run its models.

Second, the money. Anthropic has not yet disclosed a target share price, but it is widely expected to debut at a valuation of $2 trillion or more, which would make it the largest listing in history. It lost $80.7bn (£60bn) in the first half of this year, double its losses across all of 2025, on revenue of $16.2bn for the six months.

Third, it still holds $131.6bn in cash, almost all of it working capital, more than Meta and Google combined, although leasing data centres alone has cost it $321.4bn so far this year.

Fourth, its business is concentrated: 250 major enterprise customers paying at least $10m each in the past year account for a large share, and just five make up a quarter of it, although its enterprise customer base has grown from about 1,500 last year to almost 12,000.

Finally, Anthropic says AI will bring extraordinary benefits, from bespoke medical treatments and better disease detection to tailored education and more efficient power grids, and that frontier AI will help drive global economic growth.