Anthropic lifts the veil on its IPO

Like a high school student who has just sent their SAT scores and grades to colleges, Anthropic has filed some important paperwork. Yesterday, the AI company took a big step toward its much-expected IPO by filing official documents that explain its plan to list its shares on the Nasdaq stock market under the symbol ANT.

The filing gives a clearer picture of Anthropic’s finances and how it plans to manage the risks its products could pose to society, including risks that could cause serious harm.

Big numbers

Investors are worried about AI, and Anthropic’s filing gives both people who are optimistic about the company and people who doubt it plenty to think about:

• Anthropic’s revenue grew 14 times to $16.3 billion in the first half of 2026 compared with the same period a year earlier. The company expects this rapid growth to slow. Its revenue also depends less on a small number of customers than many analysts had expected.
• However, the company has not yet made a profit. It also has $417 billion in promised payments to suppliers, including companies that run data centers.

Analysts believe Anthropic could raise as much as $100 billion, beating SpaceX’s record IPO earlier this year and making Anthropic one of the world’s most valuable companies.

Things could go wrong

Anthropic says AI could bring major changes to the world, but it also used 94 pages of its filing to explain possible risks. These include lawsuits from customers who were not given early access to its most advanced models and government rules that could limit its business. The company is also worried that it may not be able to predict demand for its models accurately when deciding how much money to spend on infrastructure.

In an unusual step…Anthropic also listed risks for people who do not plan to buy its stock. It said it is worried that there may not be enough safety measures to stop people with bad intentions from using its AI to develop weapons or carry out major cyberattacks.

Anthropic said its co-founders will control important decisions, while trustees who do not have a financial interest in the company and put safety first will oversee the choice of board members. Critics said this setup would give company leaders too much control and leave shareholders with less power.

Is it a good buy? Claude, Anthropic’s AI assistant, said yes, as long as the stock is sold at a reasonable price. Its rival, OpenAI’s ChatGPT, disagreed.—SK