Anthropic, the company behind Claude, has filed to go public on Nasdaq. The filing shows huge losses, huge computing bills and risks the company admits to.
Anthropic, the maker of the Claude AI assistant, has filed paperwork to sell shares to the public. It plans to list on the Nasdaq stock exchange under the ticker symbol ANT. The filing is public, so anyone can now see how much money the company is losing and how much it has promised to spend.
The document is called an S-1. It is a prospectus, the form a company files with the Securities and Exchange Commission before its first public share sale, known as an IPO. Anthropic first filed in confidential form in June. The public version adds detail.
In the first half of 2026, Anthropic reported a net loss of $80.7 billion on $16.3 billion in revenue. The Wall Street Journal had reported, based on leaked versions, that most of the loss comes from accounting charges tied to notes the company issued for funding.
Running AI takes enormous computing power, and the filing lists what Anthropic has agreed to pay for it.
The SpaceX deal is for access to Nvidia GPUs, the specialized chips that power AI.
Anthropic wrote that AI could speed up science, lift economic growth and bring cures for major diseases within reach. It also said its behavior may differ from that of a company focused only on maximizing profit.
Its risk list includes AI models developing "forms of sentience," "AI-specific taxes" and unfavorable outcomes in copyright disputes over how its models are trained. It also warned of "catastrophic" AI-related incidents. It said no single insurer, and potentially not the insurance industry as a whole, could cover that kind of risk.
CEO Dario Amodei's pay was nearly $18 million in 2025, mostly stock and option awards. His sister Daniela Amodei, the president and a cofounder, received $16.4 million.
The material does not say when trading will begin, what the share price will be or how many shares will be sold. Those details matter most for anyone thinking of investing.
Anthropic's IPO filing shows how costly AI is to run and lists risks the company itself admits to, which matters if you use Claude or plan to invest.
Visa tested 22 AI models on realistic shopping tasks. They passed just over half, and often bought the wrong thing anyway.
Read more: Visa tested 22 AI shoppers, and most still make costly mistakesDARPA wants AI to hunt for design flaws in stock, bond, energy and cloud markets, so enemies cannot use them to cause economic damage that reaches households.
Read more: Pentagon wants AI to find weak spots in stock and energy marketsTuring Award winner Richard Sutton says more computing power alone will not make AI truly smart. Big AI companies disagree, which matters for anyone betting on AI.
Read more: AI pioneer says piling on computing power won't make machines smart
