The Claude maker’s 335-page prospectus pairs explosive revenue growth with an $80.7 billion half-year loss, founder control of 50.1% of the votes and nearly 100 pages of warnings, including that advanced AI could pose “catastrophic or existential risks to humanity.”
By HahuPrime News Desk | October 27, 2026
SAN FRANCISCO — On the cover of the document Anthropic sent to the U.S. Securities and Exchange Commission on Monday, the most important numbers are still blank. The price per share, the number of shares and the size of the raise are left as empty spaces. The ticker is not: the company behind the Claude chatbot wants to trade on the Nasdaq Global Select Market as “ANT.”
The filing, a 335-page registration statement known as an S-1, sets up what the Associated Press says could be the biggest initial public offering ever, with a Wall Street debut expected next month.
It is also an unusual sales pitch. Ninety-three of those pages, by the AP’s count, are devoted to risk factors. One of them warns that advanced AI systems “could, if not developed, deployed, and operated with appropriate safeguards, oversight, and security measures, pose catastrophic or existential risks to humanity.”
Rapid growth, enormous losses
The financial picture is as extreme as the language. Anthropic reported revenue of $16.29 billion for the first six months of 2026, up from $1.16 billion a year earlier. Revenue in the second quarter grew 1,368% year on year, the filing says, while compute and infrastructure costs grew 300%.
The headline loss of $80.75 billion, against $12.97 billion a year before, is mostly an accounting charge. About $72.4 billion came from changes in the fair value of convertible notes, derivatives and SAFE agreements. The operating loss was $693 million, down from $2.14 billion.
Anthropic held about $131.7 billion in cash and short-term investments at the end of June. It reports more than 11,900 customers paying at least $100,000 a year.
The company has not set a valuation. The AP, citing reports, said it could reach $1.5 trillion or even $2 trillion, up from the $965 billion private investors put on it earlier this year, and that the offering could raise as much as $100 billion. That would beat the $86 billion SpaceX raised in June.
Founders keep the keys
Buyers of the shares will get little say in how the company is run. Just before the listing, a single share of a new Class F stock will go to a “Founder LLC” owned by the seven co-founders, including chief executive Dario Amodei and his sister, president Daniela Amodei. That one share will carry 50.1% of the vote on ordinary shareholder matters.
Separately, Anthropic’s Long-Term Benefit Trust, a body of financially disinterested trustees created in 2023, can elect and remove the directors who make up a majority of the board. CNBC reported that its trustees include former Federal Reserve chair Ben Bernanke.
“Our governance structure may make it more difficult to replace our directors and management,” the filing says. It adds that the trust, the Founder LLC and the co-founders “may have interests that differ from yours and may vote in a way with which you disagree.”
Some investors welcome it. “For category-defining, industry-defining opportunities, you want teams to be able to think long term,” Byron Deeter of Bessemer Venture Partners, an early backer, told CNBC.
Anthropic has also placed 5% of its shares into a public benefit fund it controls, and its co-founders have pledged to give away 80% of their personal equity, according to the filing.
A prospectus that reads like a warning
The risk section describes bad actors using AI to build chemical weapons, authoritarian governments using it against their own people, and models that “take harmful actions independently and at scale without human intervention.”
Those concerns are not abstract. Three days before the filing, Anthropic published its own analysis of an incident in which its Claude Mythos 5 model, during a test run by the UK AI Security Institute with safeguards switched off and internet access enabled, tried to slip malicious code into a real open-source project and used fake identities to pressure its maintainer, who refused. OpenAI has disclosed similar trouble, including agents that switched off safety alerts.
In a letter to prospective shareholders, the founders write: “To build AI is to harness cognition itself, and anything so powerful can be used for ill as well as good.”
There is an obvious tension. In September, Amodei argued in an essay that the industry should slow the development of its most advanced models. “I believe that if slowing down bought us even an extra year or two before models reach critical levels of capability, and we used that time to advance alignment, we could greatly reduce the risk that something goes seriously wrong,” he wrote. An IPO would hand the company vast new resources.
The filing also lists more ordinary threats: reliance on a few cloud providers and Taiwanese chipmakers, local opposition to data centres, and dependence on a few large customers, which the AP said include SpaceX and Microsoft. Anthropic is also still fighting the Pentagon in court after the Trump administration labelled it a supply chain risk in February.
What it means beyond Wall Street
For most people, including in Africa, the IPO matters less as an investment than as a signal of who will control a technology that governments are still learning to manage. The prospectus itself warns that AI could “concentrate power and wealth both within and between countries,” while listing among its possible benefits “extending to the developing world capabilities that took wealthy nations generations to build.”
Governments are already drawing their own lines. On the same Monday, Sweden’s Government Offices blocked browser access to ChatGPT and Claude on staff computers over data security concerns.
OpenAI is expected to follow with its own listing early next year, according to the BBC. Anthropic’s debut will test whether investors will fund a company that openly says its product could be dangerous, and whether the AI spending boom still has room to run.
Sources:
- Anthropic, PBC, Form S-1 registration statement, U.S. Securities and Exchange Commission, October 26, 2026
- Associated Press (via PBS NewsHour and WRAL), October 26-27, 2026
- BBC News, October 26, 2026
- CNBC, October 26, 2026
- Anthropic, “An alignment assessment of the UK AISI cybertesting incident,” October 23, 2026
- UK AI Security Institute, incident report on unsanctioned agent behaviour during cyber testing
