Anthropic, the five-year-old company behind the Claude models, filed papers on Monday to sell its shares to the public for the first time, becoming the first frontier artificial intelligence lab to seek a stock market listing and opening its books at a moment of both breakneck growth and deepening unease about the cost of the AI boom.
What the filing shows
The numbers in the registration statement, filed with the Securities and Exchange Commission on Monday morning, describe a company growing faster than almost any in corporate history - and spending accordingly. Revenue in the first half of 2026 was about $16.3 billion, and revenue in the quarter ended June 30 grew 1,368 percent from a year earlier, according to the filing’s figures as reported by the Associated Press and Reuters.
The same filing reports a net loss of $80.7 billion for the first half of the year, a figure driven largely by an accounting rise in the fair value of bonds convertible into Anthropic shares rather than by operating costs alone. The prospectus also warns that the company’s revenue growth rate could decline over time and may fluctuate. The filing gauges Anthropic’s enterprise adoption by the number of customers contributing at least $100,000 in trailing-twelve-month revenue, a threshold it says characterizes organizational rather than individual use of Claude.
The debut, expected before Thanksgiving, could eclipse the record $86 billion initial public offering that SpaceX completed in June. Goldman Sachs and Morgan Stanley are listed as the offering’s banks. The number of shares and the price range have not yet been set.
A safety lab, built to stay in charge
The prospectus also lays out an unusual governance structure designed to keep the company under the control of its founders and its safety mission even after public investors come aboard. Anthropic is incorporated as a Delaware public benefit corporation whose stated purpose is the responsible development of advanced AI “for the long-term benefit of humanity.”
A Long-Term Benefit Trust of financially disinterested trustees, with expertise in AI safety, national security and public policy, holds a special class of stock that lets it elect and remove a majority of the board. A separate founder-held share carries 50.1 percent of total voting power, leaving chief executive Dario Amodei, president Daniela Amodei and their five co-founders in effective control. The filing concedes Anthropic will be a “controlled company” under Nasdaq’s rules.
The offering lands weeks after Dario Amodei published an essay arguing for slowing the development of the most advanced AI models, and the prospectus itself warns investors about the technology’s potential to cause harm - an unusual note of caution in a document meant to sell stock. Strategic business partners hold a separate class of non-voting shares that convert to common stock when sold to outsiders.
What comes next
Anthropic’s own announcement is deliberately spare: the filing is a first step, and the registration must be declared effective before any shares can be sold. A roadshow and a final price range come next. The offering is planned as a global one, with public offerings in Australia, Canada, parts of Europe, Japan, Switzerland and Britain, though the shares themselves will list only on Nasdaq. The debut is expected before the Thanksgiving holiday, according to the Associated Press and Reuters.
The listing will be read as a verdict on more than one company. Investors have spent the autumn rotating out of AI stocks, questioning whether the trillions being poured into data centers and chips will earn back their cost. A strong Anthropic debut would suggest public markets are still willing to fund the boom; a weak one would give its skeptics a number to point to.
Rival labs are watching. Anthropic’s filing is the first detailed public accounting of a frontier AI lab’s economics - revenue, losses, customer concentration and all - and it sets the benchmark against which OpenAI and others will eventually be measured when they follow.
October 27, 2026