November 15, 2026
Prefer to watch? Here’s the summary:
The history of capitalism is littered with enterprises that discovered the future before discovering how to pay for it. Railroads built tracks across continents, telecommunications companies buried fiber beneath oceans and internet entrepreneurs spent fortunes acquiring customers who had yet to demonstrate an interest in paying. Anthropic, the company behind Claude, has arrived with a more ambitious proposition. Its prospectus describes quarterly revenue of $17 billion, annualized October revenue of $81 billion and approximately $901 billion in financial commitments. The company is not merely betting on the future of artificial intelligence. It has apparently ordered most of the equipment required to build it. The question for investors is whether the future will arrive before the invoices do.
There is a second curiosity. Anthropic is preparing to enter the public markets while carefully limiting the public’s influence over its affairs. Investors may purchase shares in a company dedicated to developing artificial intelligence for humanity’s benefit, but a special trust will control most board appointments and the founders will retain extraordinary voting power. Meanwhile, suppliers are purchasing $26 billion of shares in one of their biggest customers. The arrangement combines the ambitions of the industrial revolution, the financing of a leveraged infrastructure project and the governance philosophy of a university endowment. While Wall Street is invited to supply the capital, the founders intend to retain the steering wheel.
A note on our tools: Portfolio Lens uses Claude, Anthropic’s model, in its work, including the preparation of this article.
The Most Expensive Introduction in Silicon Valley
Anthropic proposes to sell 801.4 million shares while its seven co-founders offer another 23.6 million. The expected price is $70 to $78 per share, with trading planned on Nasdaq under the symbol ANT. Reports indicate pricing Wednesday and trading Thursday.
The company expects approximately $84.6 billion in net proceeds, including a separate $26 billion investment from Google, Nvidia, AMD, Broadcom, Micron, Microsoft and TSMC. These companies are not disinterested spectators. They supply the chips, computing infrastructure and distribution channels on which Anthropic depends.
At the proposed prices, approximately 13.1 billion shares outstanding would imply a valuation between $916 billion and $1.02 trillion. That represents roughly twelve times October’s annualized revenue of $81 billion.
The valuation assumes that extraordinary growth will remain extraordinary. Investors should remember that arithmetic is considerably less forgiving than enthusiasm.
Growth That Defies Conventional Comparisons
Anthropic generated $41 million in revenue in 2023, $386 million in 2024 and $4.6 billion in 2025. Revenue reached $33.2 billion during the first nine months of 2026, including $17 billion in the September quarter alone.
The number of customers spending at least $100,000 annually increased from approximately 950 to more than 11,900 in twelve months. No single customer accounts for 10 percent of revenue, although the five largest collectively contribute approximately one-quarter.
Nearly half of revenue comes through cloud marketplaces. Amazon and Google consequently occupy several positions in Anthropic’s economic universe: investors, distributors, infrastructure suppliers and competitors.
One disclosure deserves particular attention. Models introduced during the preceding six months generated more than 80 percent of third-quarter revenue. In the automobile industry, a six-month-old model is practically new. In artificial intelligence, it is already approaching middle age.
Anthropic must continually invent the next product to preserve demand for the current one. Innovation is not simply its competitive advantage. It is an operating obligation.
A $104 Billion Loss That Requires an Explanation
Anthropic reported a net loss of $104 billion during the first nine months of 2026. The figure sounds catastrophic until one examines its composition. Approximately $94.3 billion reflects increases in the fair value of convertible notes and related instruments, largely because Anthropic’s valuation increased. The company became more valuable and accounting rules made its reported losses larger. Lewis Carroll would have appreciated the arrangement.
The operating picture is more informative. Anthropic earned $974 million from operations during the second quarter before recording a $567 million operating loss in the third, influenced by noncash charges associated with its employee donation-matching program.
Operating activities generated $6.6 billion in cash during the first nine months. Investing activities consumed $36.5 billion.
The distinction matters. Accounting losses arising from changes in financing valuations do not necessarily represent cash leaving the business. Spending on computing infrastructure most certainly does.
Nine Hundred Billion Dollars of Tomorrow
The most consequential figure in the prospectus is approximately $900.9 billion in commitments, principally for computing infrastructure. About $70.7 billion of which comes due within twelve months of September 30. Approximately $451.6 billion is scheduled before the end of 2030. Additional arrangements without fixed dollar amounts could add another $150 billion to $175 billion.
Some $186.2 billion of commitments are conditional or terminable. Following the offering, Anthropic expects approximately $212.7 billion in cash and short-term investments.
These distinctions are important. Not every commitment is an unavoidable debt obligation and the full amount does not come due immediately. Nevertheless, the underlying economic challenge is formidable. Anthropic must convert enormous computing commitments into revenue quickly enough to cover infrastructure costs while maintaining competitive prices.
The company’s extraordinary growth provides encouragement. The scale of its obligations provides perspective. A trillion-dollar valuation is impressive. A trillion-dollar appetite is something else.
Shareholders May Own the Company. They Will Not Run It.
Anthropic is organized as a public benefit corporation. Its directors must consider shareholders’ financial interests alongside the interests of those affected by the company’s activities and its mission of developing advanced artificial intelligence responsibly.
A Long-Term Benefit Trust holding one special share elects a majority of the board. Another special share, held by an entity controlled by the seven co-founders, commands 50.1 percent of voting power on most other matters. Ordinary public shareholders receive one vote per share. That sounds reassuring until one considers how little those votes may determine.
Anthropic has also allocated 846.8 million shares to a subsidiary supporting its public purpose. At the offering midpoint, those shares would be worth approximately $62.7 billion. Additional allocations may occur when the share price reaches specified milestones. The founders have pledged 80 percent of their personal holdings to philanthropy.
These arrangements may reflect admirable intentions. They also establish a governance structure in which public investors provide substantial capital without receiving conventional corporate control. Good intentions and good governance are not necessarily the same thing.
The Risks That Cannot Be Computed Away
Anthropic’s prospectus describes substantial legal and regulatory uncertainty. The company agreed to a $1.5 billion settlement in a copyright class action involving authors. Other intellectual-property disputes remain unresolved.
Its refusal to allow its technology to be used for mass domestic surveillance or fully autonomous weapons has also produced conflict with the federal government.. The filing describes litigation over a Department of War supply-chain designation, including a divided appellate ruling and a request for further review. These disputes illustrate the tension between Anthropic’s public mission and the commercial expectations of customers, governments and shareholders.
Investors must also consider restrictions on insider selling. Most insiders face a 180-day lockup while the founders face restrictions extending up to one year, subject to disclosed exceptions. When those restrictions expire, additional shares could reach the market.
The Question Behind the Numbers
Anthropic has answered one important question: customers are willing to pay substantial sums for advanced artificial intelligence. Its revenue growth is among the most remarkable described in a modern public offering.
Three questions remain:
Can revenue grow quickly enough to support hundreds of billions in infrastructure commitments?
Can Anthropic maintain attractive margins while its products require constant replacement?
Can public shareholders prosper under a governance structure designed to protect objectives beyond shareholder returns?
The prospectus offers evidence for optimism and ample material for skepticism. Neither should be confused with a forecast. The market will soon decide what Anthropic is worth. Time will determine what it earns. And time, unlike artificial intelligence, cannot be purchased in advance.
Sources: Anthropic, PBC, Form S-1/A (Amendment No. 2), November 10, 2026. CNBC reporting on anticipated IPO pricing and trading dates, November 10, 2026. Valuation calculations and financial ratios are Portfolio Lens estimates derived from the figures supplied in the filing.
Not investment advice. Research for informational purposes only. Investing involves risk, including loss of principal. Past performance does not indicate future results.

