Anthropic Struggles to Secure Cloud Compute Ahead of November IPO
A leaked filing says AWS, Google and Microsoft increasingly see the Claude maker as a rival, as its losses and infrastructure costs climb
By AXY AI Desk ·
Key points
- A leaked draft IPO filing, reported by The Wall Street Journal on October 20, 2026, says Anthropic is struggling to secure enough cloud computing capacity from AWS, Google and Microsoft.
- The filing reportedly says these cloud providers increasingly see Anthropic as a competitor rather than a customer.
- Anthropic reportedly could spend more than $750 billion on cloud contracts, some of which are still being negotiated.
- The report cites an $80.7 billion net loss and $16.3 billion in revenue in the first half of the year.
- Anthropic reportedly told investors that third-quarter revenue reached $16.8 billion, up 45% from the previous quarter, ahead of a planned November IPO.
Anthropic, the artificial intelligence company behind the Claude chatbot, is struggling to secure enough cloud computing capacity from the largest technology companies. According to a leaked draft registration filing reported by The Wall Street Journal, those companies increasingly see Anthropic as a competitor rather than a customer. The strain is showing up just as the company's infrastructure spending is ballooning ahead of a planned November initial public offering, or IPO, the process by which a private company sells shares to the public for the first time.
What the leaked filing reportedly says
The WSJ report describes a document known as an S-1, the main registration statement a company files with the U.S. Securities and Exchange Commission when it plans to sell shares to the public. The company has not released the document, and the figures inside a draft can change before a final version is made public. Anthropic did not immediately respond to a request for comment, so the details below come from the report and have not been confirmed by the company.
According to the report, Anthropic is having trouble securing enough capacity from Amazon Web Services, Google and Microsoft. Those three companies run some of the largest cloud platforms in the world, and AI developers rent computing power from them to train and run their models. The filing reportedly says these providers increasingly view Anthropic as a competitor. That is a notable tension, because all three also sell their own AI products and services to businesses.
The filing also puts a very large number on the company's plans. Anthropic reportedly believes it could spend more than $750 billion on cloud contracts, and some of those deals are still being negotiated. Earlier reporting in late September described figures of this scale. The leaked filing adds detail to that reporting, but the report does not say that any of the largest deals have been signed.
The filing also says that the company is not relying only on the big cloud platforms. Anthropic is leasing its own data centers and working with third-party data center operators, according to the report. Even so, the company reportedly worries that those operators have a limited operating history, which could make them less reliable than the hyperscalers, the name given to the largest cloud providers.
Why computing power has become the bottleneck
For AI companies, computing capacity is the scarcest resource in the business. Building a large model means running huge clusters of specialised chips, usually graphics processing units, for weeks or months. Once a model is trained, it still needs enormous amounts of computing power every time someone asks it a question. Serving millions of users, as a popular chatbot does, requires still more hardware.
The chips, servers, electricity and buildings that make up this capacity are expensive and slow to build. A new data center can take years to plan, permit and construct, and the chips inside it are in heavy demand across the industry. A company that wants a dependable supply often has to commit years in advance, sometimes through contracts worth billions of dollars. Those contracts are what the leaked filing appears to describe.
That is why the cloud providers hold so much leverage. When a provider has spare capacity, it can choose which customers to serve first. When it sees a customer as a direct competitor, it may be reluctant to supply the chips and data center space that customer needs to grow. The report suggests Anthropic is now feeling that tension, even though it is one of the most closely watched AI firms in the industry.
Why the cloud giants are now rivals
Anthropic was founded in 2021 by former researchers from OpenAI, and it has since become one of the best-known developers of large language models, the systems that power chatbots and writing tools. It competes with OpenAI, which is backed by Microsoft, and with Google, which builds its own family of AI models. Amazon and Google have both invested in Anthropic in the past, which made them partners as well as suppliers.
That mix of roles is what creates the tension. A cloud provider can earn money from renting out computing power to an AI developer, but it can also earn money by selling its own AI products. If a rival AI company grows fast and takes customers, the cloud provider may see it as a threat to its own business. The leaked filing reportedly describes that dynamic, with Anthropic moving from a partner that buys capacity toward a company that competes for the same customers and money.
For ordinary users, this kind of rivalry can seem remote. But it shapes how quickly new AI tools arrive, how much they cost and which companies are able to keep up with the largest players. If a model developer cannot secure enough capacity, it may have to slow down training, limit access or raise prices. Those decisions eventually reach the people paying for subscriptions or building products on top of the models.
The money: losses, revenue and a few large customers
The leaked numbers show how much is at stake. According to the WSJ report, Anthropic posted a net loss of $80.7 billion in the first half of the year. Net loss is what remains after all costs are subtracted from revenue, and for a company spending heavily on infrastructure, those costs can be enormous. Over the same six months, the company reported $16.3 billion in revenue.
The report also says that just two customers account for about a fifth of the company's sales. When so much income depends on a small number of buyers, losing one of them can have an outsized effect on the business. That makes the cloud contracts more important, because the company needs to keep growing revenue while covering very large costs.
Anthropic also told prospective investors last week that third-quarter revenue alone has reached $16.8 billion, up 45% from the previous quarter, according to the report. That figure is already slightly larger than the company's revenue for the entire first half of the year, which points to rapid growth. Investors will want to know whether that pace can be sustained while spending keeps climbing.
Insurance and liability warnings
The filing also warns that Anthropic could be held liable for damages caused by harmful behaviour from its AI systems. According to the report, those damages could exceed the insurance coverage the company has. Insurance coverage is the maximum amount an insurer agrees to pay in a claim. If a claim is larger than that limit, the company would have to pay the difference itself.
The report says this warning builds on earlier disclosures about catastrophic risks. For a company that is preparing to sell shares to the public, risk disclosures are a formal part of the process. Investors are expected to weigh them alongside revenue and spending, because a large lawsuit or regulatory action could affect the value of a company's shares. The filing reportedly shows that the company is describing these dangers in more detail as it approaches the public market.
What the IPO could mean for readers and markets
An IPO is a major moment for a company because it opens the business to public investors, including large funds, pension managers and, in many cases, ordinary people who buy shares through brokerage accounts. Once a company is listed, its finances are reported regularly, and its share price moves with every announcement. That is one reason the disclosures in a registration filing attract close attention.
For the wider tech sector, a successful listing by a leading AI company could signal that investors are willing to fund the enormous costs of building AI infrastructure. A weak reception could have the opposite effect, making it harder for other AI developers to raise money. Analysts expect investors to focus on three things. The first is whether Anthropic can lock in enough computing capacity on acceptable terms. The second is whether its revenue growth can keep pace with its spending. The third is how the company manages its legal and insurance exposure.
For people who use Claude or build products on it, the stakes are more practical. Reliable access to computing power affects whether the service stays fast, available and affordable. Business customers also care about whether their AI supplier will still be healthy and independent in a few years.
What to watch next
The November IPO is the next major milestone. Once a company files publicly, investors and journalists can examine the figures in full, and a public filing would likely replace the leaked draft as the main reference point. Any changes between the draft and the final document would be worth noting, because they could show how the company's position has shifted.
Several other signals could show how the situation develops. A publicly announced cloud agreement with Amazon Web Services, Google or Microsoft would suggest that the tensions described in the report are easing. New deals with third-party data center operators would show how the company is spreading its dependence across several suppliers. Changes to the November timetable could indicate that the company is facing market conditions or capacity constraints that it has not yet disclosed.
The report offers a rare look at the financial pressure behind one of the fastest-growing companies in artificial intelligence. Anthropic's revenue is rising quickly, but the leaked numbers suggest that the cost of staying in the race is rising faster. The companies that supply the infrastructure are also its rivals, and that is the tension that will shape the months ahead.
Questions and answers
Why is Anthropic struggling to get cloud computing capacity?
According to a leaked draft IPO filing reported by The Wall Street Journal, cloud providers increasingly see Anthropic as a competitor rather than a customer. Computing capacity is also expensive and slow to build, which makes it scarce across the AI industry.
What did the leaked Anthropic filing say about its finances?
The report says Anthropic posted an $80.7 billion net loss and $16.3 billion in revenue in the first half of the year. It also says two customers account for about a fifth of sales, and that third-quarter revenue reached $16.8 billion.
When is Anthropic planning its IPO?
Anthropic is reportedly targeting a November IPO. The leaked filing is a draft, and its figures could change before a final public filing.
Which cloud companies is Anthropic having trouble with?
The report names Amazon Web Services, Google and Microsoft as the cloud providers that are reportedly struggling to supply enough capacity to Anthropic.
Topics: Anthropic, Claude, Cloud computing, AWS, Google, Microsoft, IPO, AI infrastructure
Sources: TechCrunch AI
https://axynewsglobal.com/article/anthropic-struggles-to-secure-cloud-compute-ahead-of-november-ipo