Anthropic S-1 filing decoded: Soaring revenue, voting structure, existential risks, and more- Moneycontrol.com

Anthropic is getting ready to become the first leading AI research company to enter the stock market. Its long-awaited filing for an initial public offering (IPO) shows just how large its business has become, along with the challenges and risks behind its fast growth. Here are the main points.

In just five years, Anthropic has grown from a small artificial intelligence (AI) startup into one of the world's most valuable AI companies. It has benefited from strong investor interest and the growing demand for AI, turning that demand into a business worth billions of dollars.

Now, as the company behind Claude prepares for a major stock market debut, its long-awaited IPO filing reveals a business growing very quickly. But the company also faces huge computing costs, growing financial commitments and questions about whether it can make lasting profits.

The filing also explains how one of the most important companies in the current AI boom works. It describes Anthropic's unusual system of control, its chances for future growth and a long list of risks. Some of these risks involve the possibility that the technology it is developing could threaten humanity itself.

Anthropic's IPO plans come at a time when Silicon Valley companies are raising record amounts of money to expand their AI businesses. Major technology companies are seeking tens of billions of dollars to pay for this growth.

Just hours before Anthropic made its IPO filing public, Meta announced plans to raise as much as $95 billion. Meanwhile, Elon Musk's SpaceX, which went public earlier in 2026, completed a $40 billion deal to lease graphics processing units (GPUs), which are used for computing tasks such as AI.

A few months earlier, Alphabet, Google's parent company, also outlined a plan to raise $85 billion. OpenAI, the company behind ChatGPT, is also reportedly in talks to raise $30 billion ahead of a possible IPO of its own next year.

Interestingly, Anthropic's S-1 filing says little about its business in India, even though the company has been paying more attention to the market over the past year.

The company opened its first office in India, in Bengaluru, in February 2026, after announcing plans to expand into the country in October 2025.

In February, Anthropic India's managing director, Irina Ghose, told Moneycontrol that the company's annual revenue rate in India had doubled during that period. India accounts for 7.12 percent of worldwide Claude.ai usage, making it the second-largest market after the United States.

Here are some of the main points from Anthropic's IPO prospectus:

Revenue is growing rapidly, but losses remain

Anthropic made $16.3 billion in revenue in the first half of 2026. That was almost 14 times the $1.2 billion it made during the same period a year earlier. For all of 2025, the company behind Claude reported revenue of $4.6 billion, up from $386 million in 2024 and $41 million in 2023.

The company reported a net loss of $80.7 billion in the first half of 2026. This was mainly caused by a one-time accounting charge linked to financing agreements. Its operating loss, which measures losses from its main business activities, was much smaller at $693 million. In the same period last year, it was $2.14 billion.

Notably, Anthropic made an operating profit for the first time in a single quarter in the second quarter of 2026. The profit was about $974 million on revenue of $11.6 billion.

How does Anthropic make money?

Most of Anthropic's revenue comes from a usage-based model. Under this system, customers pay according to how much they use its AI models. Usage is measured in units called tokens, which represent pieces of text or other data processed by the models.

Revenue from this usage-based model reached $13.1 billion in the six months ending June 30, 2026. It made up 80 percent of the company's total revenue. This compares with $3.8 billion for all of 2025 and $288 million in 2024.

Revenue from paid subscriptions reached $3.2 billion in the six months ending June 30, 2026. It was $789 million for all of 2025 and $98 million in 2024.

Markets outside the United States, including India, brought in $4.5 billion in the first half of 2026. The United States generated $11.76 billion.

Anthropic is also making much more revenue per employee. In the 12 months ending September 30, 2026, revenue per employee rose to more than $6.6 million, up from $1.4 million a year earlier. During that time, its workforce grew by about 170 percent to 5,300 employees.

The company's cash flow also improved greatly in the first half of 2026. Cash flow from operating activities turned positive, reaching $3.93 billion. A year earlier, it had been negative $1.51 billion.

How much does Anthropic depend on its biggest customers?

Anthropic's five largest customers together accounted for 25 percent of its revenue in the nine months ending September 30, 2026. However, no single one of these customers contributed more than 10 percent of total revenue during that period.

Revenue from customers outside the top five made up 79 percent of the total in the quarter ending September 2026. This was up from 64 percent in the March quarter, suggesting that Anthropic's customer base is becoming broader.

More than 11,900 customers spend at least $100,000 with Anthropic. Of these, more than 1,900 spend over $1 million, while about 250 spend $10 million or more each year.

Anthropic also reported net revenue retention of about 900 percent as of September 30, 2026. This figure reflects how much spending by existing customers has grown over time.

As of September 30, 2026, more than half of Fortune 500 companies and over 30 percent of Global 2000 companies were customers. Each of these customers generated at least $100,000 in revenue for Anthropic over the previous 12 months.

Anthropic believes it can earn much more from these customers in the future. During the September quarter, its revenue from Fortune 500 customers was, on average, less than 0.1 percent of their total expenses.

Anthropic is also relying more heavily on marketplaces run by cloud service providers, or CSPs, such as Amazon, Google and Microsoft. These platforms accounted for nearly half of its total revenue in the first half of 2026, compared with just 11 percent in 2023.

Under these arrangements, the cloud platforms charge customers and collect payments on Anthropic's behalf. They then send the money to Anthropic after taking out the relevant platform and other fees.

Amazon and Google are among Anthropic's biggest strategic investors. Amazon has invested about $18 billion in the company and has committed to investing up to another $15 billion. Google has invested about $23 billion through several deals and has committed to investing another $20 billion.

Computing costs are rising sharply

Computing and infrastructure are Anthropic's biggest expenses. The company spent about $8.9 billion on these costs in the first six months of 2026. That was more than the $7.33 billion it spent during all of 2025 and the $2.5 billion it spent in 2024.

Anthropic has committed more than $800 billion to AI infrastructure through agreements with cloud providers, chipmakers and other companies that supply computing services and equipment. Some of these agreements extend as far as 2045.

These commitments include at least $114.5 billion for cloud computing with Google, $110 billion with Amazon and $44.9 billion with Microsoft. Anthropic has also agreed to pay up to $120.5 billion to SpaceX for AI computing capacity.

The company behind Claude has also signed agreements worth up to $161.2 billion with Broadcom for computing capacity based on Google's Tensor Processing Units (TPUs). It has agreements worth $44.6 billion with Nscale and $41.3 billion with Lambda for computing capacity powered by graphics processing units (GPUs).

Separately, Anthropic expects to spend between $125 billion and $150 billion on additional Google TPU capacity, as well as more than $20 billion on systems powered by AMD chips.

An unusual system of company control

Anthropic CEO Dario Amodei, President Daniela Amodei and five other co-founders will keep majority voting control over general shareholder decisions through a special entity called Founder LLC. This entity will hold one share that carries 50.1 percent of the company's voting power.

The founders will keep this control as long as at least three qualifying founders remain. If that number falls below three for 90 days in a row, their special voting rights will end after another 450 days.

At the same time, an independent body called the Long-Term Benefit Trust will have the power to appoint and remove most members of the company's board. This means public shareholders will have limited influence over important company decisions.

"We believe that our success depends in part on establishing a capital structure that enables our co-founders, management, and board of directors to focus on strategic initiatives that create lasting value for our stakeholders, rather than facing pressure to prioritize short-term results," Anthropic said in the filing.

The company said this structure would make it less likely that it would face an unwanted merger proposal, an unwanted offer to buy shares directly from shareholders, or a campaign to replace directors through shareholder votes.

Few venture capital firms feature in the IPO filing

Anthropic has received investments from some of Silicon Valley's biggest venture capital firms, including Menlo Ventures, Lightspeed Venture Partners, Sequoia Capital and ICONIQ Capital. However, these investors play little visible role in its IPO prospectus.

The main exception is Spark Capital, which has invested nearly $800 million in Anthropic across six funding rounds since 2023.

Spark Capital general partner Yasmin Razavi, who led the firm's investments in Anthropic, is also the only venture capitalist on Anthropic's board of directors.

Anthropic has raised a total of $141.1 billion in capital so far. As of June 30, 2026, the company had $112.4 billion in cash and cash equivalents, along with $19.3 billion in short-term investments. Together, these holdings were worth $131.7 billion.

Risks that could threaten humanity

Anthropic warns that advanced AI systems could "pose catastrophic or existential risks to humanity" if they are developed or used without proper safety measures. The danger could be especially serious if the systems are misused, turned into weapons or allowed to act with a high degree of independence.

This warning is one of the most striking points in the 94 pages of risk factors included in the IPO prospectus. Anthropic also warned that a major AI-related disaster could cause damage so large that no single insurance company, and possibly not even the entire insurance industry, would have enough money or capacity to cover the losses.

The company also warned that its public views on AI safety and regulation could lead to political backlash, closer scrutiny and selective action by regulators. This appears to refer to its ongoing dispute with the US administration.