Anthropic’s IPO, Can It Succeed?
Anthropic, the developer of the chatbot Claude, publicly filed an S-1 with the U.S. Securities and Exchange Commission on October 26. An S-1 is the formal document a private company files before selling shares to the public for the first time, disclosing its business, finances and risks in detail. With the filing, Anthropic applied for a Nasdaq listing under the ticker ANT, the short symbol under which its shares would trade. Goldman Sachs and Morgan Stanley are the lead underwriters, the investment banks arranging the listing and selling the shares.
Much is still unsettled. The price range, the band within which shares would be sold, along with the share count and total proceeds, are left blank in the filing. The company has not announced a listing date. Reports point to a roadshow starting in the week of November 9, followed by a listing in mid-November, before Thanksgiving. A roadshow is the process in which a company meets institutional investors ahead of a listing to explain its business and its shares. The public filing came about five months after Anthropic submitted a confidential draft on June 1.
The reported scale is large. In a May 2026 funding round of 65 billion dollars, Anthropic’s valuation, the value placed on the whole company, was confirmed at 965 billion dollars. Expectations reported for the listing run to a valuation of 1.5 trillion to 2 trillion dollars and a raise of up to 100 billion dollars. If achieved, it would be the largest initial public offering on record, surpassing SpaceX’s June listing. These figures are reported expectations, not announcements by the company, and that distinction matters.
Revenue Soared, and So Did the Loss
The revenue curve in the filing is steep. Revenue was 41 million dollars in 2023, 386 million dollars in 2024 and 4.585 billion dollars in 2025. Revenue for the first half of 2026 reached 16.286 billion dollars, about 14 times the 1.163 billion dollars of the same period a year earlier. Revenue for the second quarter of 2026 alone was 11.56 billion dollars.
The net loss for the same six months was 80.748 billion dollars, but reading that as cash the company burned would be wrong. Most of it, 72.388 billion dollars, came from changes in the fair value of convertible notes, derivatives and SAFEs. Convertible notes and SAFEs are forms of investment that later convert into shares. When a company’s value rises, the book value of the shares those investors will later receive rises with it, and accounting rules record that increase as a loss. No cash left the company. It is a paper loss, known as a non-cash loss.
The operating business tells a different story. Operating loss, the core business’s revenue minus its core costs, was 693 million dollars for the first half. In the second quarter, Anthropic recorded its first quarterly operating profit, 974 million dollars on 11.56 billion dollars of revenue. Operating profit is what remains from the core business alone, before interest, taxes and accounting valuations. That compares with an operating loss of 8.06 billion dollars for all of 2025.
The customer base grew alongside revenue. As of September 30, 2026, more than 11,900 customers spent at least 100,000 dollars a year, up from just over 950 a year earlier. About 250 customers spent 10 million dollars or more. As of June 30, the company held 112.4 billion dollars in cash, 19.3 billion dollars in short-term investments and 3.2 billion dollars available under a credit facility.
The Money Is Already Committed
The listing is less a sale of a company newly making money than a sale of a company already carrying enormous spending commitments. The computing power needed to train and run AI models is called compute. Anthropic has contracted in advance for cloud and chip capacity and promised to pay for it, with figures reported from filing summaries at about 114.5 billion dollars to Google, 110 billion dollars to Amazon, 44.9 billion dollars to Microsoft and 161.2 billion dollars in Broadcom leases. Such commitments often must be paid even if the capacity goes unused, so the burden remains if revenue slows.
That structure frames the question investors must answer. It is whether revenue can keep growing faster than the commitments. Reports already point to slowing quarterly growth. Reuters reported on October 15 that investors were discussing a valuation around 1.5 trillion dollars rather than 2 trillion dollars, citing quarterly growth slowing from about 147 percent to about 45 percent. That slowing is reported investor discussion, not a confirmed fact, and it should be read that way.
Who Controls the Company
Control will not pass to public shareholders after the listing. The founders will keep 50.1 percent of the votes. Voting rights are the rights to cast ballots on decisions at shareholder meetings. A special arrangement called a Long-Term Benefit Trust adds to that control. The trust was created to protect the company’s mission of safe AI, and it elects a majority of the board through a class of shares known as Class T stock. A founder limited liability company also holds a single share of Class F stock. Classes of shares are shares designed with different voting rights or other rights. The result is a structure in which ordinary investors buy shares but do not directly elect directors. Critics call that weak investor protection, while supporters say it protects the founders’ long-term vision.
What the Risks Are
The filing lists risks at length. They include the possibility of AI incidents at a catastrophic level, dependence on particular suppliers and regulatory change. It also records the company’s exclusion from some Pentagon contracts. Recent reporting has raised allegations of unauthorized access by Claude to Georgia voter data, and the filing refers to such security issues. Those allegations are reported claims, not established facts, and the line should be kept.
Anthropic’s strength is its position with business customers. In Ramp’s September 2026 index, which tracks company spending, 43.8 percent of U.S. businesses on Ramp paid for Anthropic products, ahead of OpenAI’s 40.6 percent. Menlo Ventures estimated in December 2025 that Anthropic held about 40 percent of enterprise spending on AI models and about 54 percent of spending on AI for coding. OpenAI leads with consumers through ChatGPT. That Anthropic is a business-market company is the starting point of the listing story.
What Is Still Undecided
Readers should keep confirmed facts separate from reported expectations. Confirmed are the S-1 filing and the Nasdaq application, the financial figures through the first half, and the governance structure. Not confirmed are the price range, the share count, total proceeds, the listing date and the valuation. Third-quarter revenue of 16.8 billion dollars is also not in the filing. The filing’s financial figures run through June 30, and the third-quarter figure was reported from investor communications. The same applies to discussions of growth rates and valuation.
Why the listing matters now is also clear. Other AI companies have pushed back their listings. Reuters reported that data center operator Switch and AI infrastructure firm Nscale delayed theirs. OpenAI filed confidentially but pushed its listing to 2027. Anthropic is first to the test of how public markets value a leading AI lab. When listing-day results arrive, the price and first-day reaction deserve a separate follow-up article.
