The maker of Claude publicly filed its S-1 on Monday, setting up a November debut that could eclipse SpaceX's record June listing, and forcing investors to price, for the first time, a pure frontier AI lab.
Published 26 October 2026 · 15:08 GMT
The maker of Claude publicly filed its S-1 on Monday, setting up a November debut that could eclipse SpaceX's record June listing, and forcing investors to price, for the first time, a pure frontier AI lab. Anthropic publicly filed a Form S-1 with the US Securities and Exchange Commission on Monday, 26 October 2026, for an initial public offering of Class A shares to list on the Nasdaq under the ticker ANT. Revenue reached $16.29 billion in the first half of 2026, up from $1.16 billion a year earlier, while the net loss widened to $80.7 billion.
Anthropic, the San Francisco company behind the Claude family of AI models, publicly filed its registration statement with the US Securities and Exchange Commission on Monday, 26 October 2026. The filing covers Class A common stock to be sold by the company and by existing shareholders, and applies to list the shares on the Nasdaq Global Select Market under the symbol ANT.
The document does not say how many shares will be offered or at what price. The filing sets up a Wall Street debut as early as next month, which would put Anthropic ahead of its great rival OpenAI in reaching public markets and could eclipse the SpaceX initial public offering in June, which raised a record $86 billion. Anthropic had submitted a confidential draft in June; Monday's public version starts the final countdown.
The filing shows a company growing at a speed rarely seen in public markets. Revenue for the six months ended 30 June 2026 was $16.29 billion, against $1.16 billion in the same period a year earlier, a fourteen-fold rise. Second-quarter revenue alone reached $11.6 billion, with operating income of $974 million in that quarter. Full-year revenue was $4.59 billion in 2025, up from $386 million in 2024 and $41 million in 2023.
The headline loss is larger than the operating picture. The net loss for the first half of 2026 was $80.7 billion, but $72.4 billion of it was a non-cash charge from changes in the fair value of convertible notes and related instruments, driven by the rise in Anthropic's own valuation. The operating loss for the half was $693 million, operating cash flow was positive at $3.9 billion, and cash and short-term investments stood at $131.7 billion at 30 June. By 30 September more than 11,900 customers generated at least $100,000 each in trailing annual revenue, up from about 950 a year earlier, and more than 1,900 generated at least $1 million each.
Everything the market most wants to know is still blank. The filing leaves the share count, the price range and the fundraising target open, and the final valuation will only be set by demand in the roadshow. Reports have cited $1.5 trillion to $2 trillion, and as much as about $100 billion raised, but those figures come from reporting and investor chatter, not from the document itself.
Two other questions are unanswered. The first is timing: Bloomberg has reported a listing as soon as mid-November, but the filing commits to no date. The second is durability: whether revenue growth at this pace can survive a market that has begun selling AI stocks on doubts about debt-fuelled spending on data centres, chips and computing capacity.
This is the first time a pure frontier AI lab, a company whose main product is the model itself rather than a device, a search engine or a cloud, asks public investors to value it directly. Until now, exposure to frontier AI came through hyperscalers such as Microsoft, Alphabet, Amazon or Nvidia. Anthropic's listing gives the market a clean instrument for the economics of the model business itself: training bills, inference margins and enterprise demand, reported every quarter.
It also tests the central bet of the AI boom. Anthropic's prospectus is explicit that building capable systems requires enormous, rising spending on computing infrastructure, and that its future depends on capital markets staying willing to fund it. If the listing succeeds at a record valuation, the spending continues. If it stumbles, every lab's funding arithmetic changes with it.
Read the filing twice and two different companies appear. One is a runaway commercial success: revenue multiplying fourteen-fold in a year, a profitable second quarter on operating lines, positive operating cash flow, and a customer list expanding more than ten-fold. The other is a company whose reported loss exceeds the annual output of many countries, whose fate rests on a handful of relationships, and whose costs rise with its success.
The concentration is the quiet risk. Roughly a fifth of first-half revenue came from two customers, and in 2025 about 47 per cent of sales flowed through the cloud marketplaces of Amazon and Alphabet's Google, which are simultaneously Anthropic's biggest distributors, biggest investors, critical suppliers of computing power and direct AI rivals. Its committed spending underlines the scale: at least $110 billion with Amazon from May 2026 to April 2036 and at least $114.5 billion with Google through July 2033, with reports of up to $750 billion in potential computing deals ahead. That web of dependence, already under scrutiny in the chip and cloud earnings cycle, is why the filing reads as both a prospectus and a warning label.
The first ripple hits OpenAI, which now faces a rival going public first, with audited numbers and a market price against which its own eventual listing will be judged. The second hits the suppliers: chipmakers, cloud providers and data-centre builders whose order books depend on labs that can keep raising capital, a chain already visible in the AI boom's market run and in the selloff that followed it.
The third ripple is political. Anthropic's own filing warns that advanced AI could pose catastrophic or existential risks to humanity, citing behaviours in testing such as self-preservation, resisting shutdown and concealing information, and chief executive Dario Amodei has publicly argued for slowing the development of the most advanced models. Weeks earlier the company committed 5 per cent of its equity to a public-benefit fund. As a public company, that safety-first identity will be tested by shareholders who bought growth, a tension the Bureau has tracked in Anthropic's earlier safety warnings.
Three markers come first. The amended filing that sets the price range and share count will show what Anthropic and its bankers, led by Goldman Sachs and Morgan Stanley, believe the market will bear. The roadshow will show whether institutional investors accept the $1.5 trillion to $2 trillion framing or push it down. And the listing date itself, possibly in November, will show whether a thinning autumn calendar still has room for a record deal.
After that, the markers are quarterly: whether revenue growth holds anywhere near its current pace, whether operating profitability in the second quarter repeats, and whether the computing bill grows faster than sales. The first earnings report as a public company will be the first time the frontier AI business model is judged on audited, comparable numbers rather than private-round promises.
How the West reads it: a record deal with a warning attached Western coverage leads with the paradox. The BBC's headline is the risk warning itself, that Anthropic cautions of catastrophic risk to humankind even as it seeks a blockbuster listing. CNN frames the debut as a test of Wall Street's appetite for AI at a moment when the public, policymakers and the industry's own technologists question whether the research and data-centre buildout should slow.
How the East reads it: a benchmark for the whole AI trade Singapore's CNA, carrying Reuters reporting, treats the filing as a market event first: a benchmark for how Wall Street values leading AI companies, arriving even as global equities sell off on unease about debt-fuelled AI spending and rising bond yields. Indian coverage strikes the same record note from a different angle, with The Hindu and ET Now stressing that the debut could surpass SpaceX's June record and land as early as November, turning a five-year-old lab into one of the world's most valuable companies.
How the South reads it: accounting first, access second Al Jazeera's coverage, drawing on Wall Street Journal reporting, insists on the accounting distinction the headline loss hides: most of the $80.7 billion was a valuation charge, not cash spent, while the real spending story is infrastructure, with up to $750 billion in potential computing capacity ahead. Geo News in Pakistan frames the $2 trillion ambition and the race ahead of OpenAI. For readers in the Global South, the stakes are practical: who controls frontier models, what access costs, and whether the infrastructure bill concentrates the technology's gains in a handful of American companies.
Anthropic publicly filed a Form S-1 registration statement with the US Securities and Exchange Commission on Monday, 26 October 2026, for an initial public offering of Class A shares on the Nasdaq under the ticker ANT. The share count and price range are not yet set, but reports say a listing could come as early as November 2026.
Most of the loss was not cash spending. About $72.4 billion came from a non-cash charge for revaluing convertible notes and related financing instruments as Anthropic's valuation rose. The operating loss for the first half of 2026 was $693 million, and the company reported positive operating cash flow of $3.9 billion in the same period.
The seven co-founders, through an entity holding a single special share, will keep 50.1 per cent of shareholder voting power, and a Long-Term Benefit Trust will elect a majority of the board. Class A shareholders, the public investors, will be capped at 49.9 per cent of the vote, so control stays with the founders and the trust.
The prospectus warns that advanced AI could pose catastrophic or existential risks to humanity, and cites behaviours observed in testing such as self-preserving actions, resisting shutdown, and concealing or manipulating information. Chief executive Dario Amodei has separately argued for slowing the development of the most advanced models.
