Anthropic faces cloud capacity constraints despite vast contracts
Anthropic is struggling to secure sufficient cloud capacity from Amazon Web Services, Google and Microsoft despite contemplating more than $750 billion in cloud contracts, some of which remain under negotiation. The disclosure appeared in an S-1 filing leaked to The Wall Street Journal, adding detail to figures previously reported by Reuters and the Financial Times.
The Claude developer is also leasing its own data centres and working with third-party data-centre operators. However, the filing warned that these operators may be less dependable than hyperscale providers because they have a limited operating history.
Cloud suppliers are also AI competitors
Anthropic’s dependence on major cloud platforms is becoming more complicated as AWS, Google and Microsoft increasingly view the company as a competitor as well as a customer. The filing portrays capacity procurement as an operational constraint even alongside exceptionally large planned commitments.
The scale of the relationship with AWS is illustrated by Anthropic’s $5 billion Amazon investment and AWS commitment as Anthropic expands the infrastructure needed to run and develop its AI models. The new filing indicates that contractual spending alone does not guarantee that the company can obtain all the computing capacity it needs.
Losses, revenue concentration and liability exposure
Anthropic reported an $80.7 billion net loss for the first half of the year while generating $16.3 billion in revenue over the same period. The filing said two customers accounted for roughly one-fifth of revenue, highlighting customer concentration alongside the company’s infrastructure commitments.
Last week, Anthropic told prospective investors that third-quarter revenue reached $16.8 billion, a 45% increase from the prior quarter. The company is reportedly targeting an initial public offering in November, while it did not immediately respond to TechCrunch’s request for comment.
The filing also expanded on previously reported catastrophic-risk disclosures. Anthropic warned that it could face liability for damages caused by harmful AI behaviour in amounts exceeding its available insurance coverage.
What businesses should take from the filing
For businesses building on AI platforms, the disclosure shows that access to compute, the reliability of alternative infrastructure providers, customer concentration and liability provisions can affect operations at the same time. Procurement plans should therefore examine capacity terms and supplier dependencies alongside model performance and commercial pricing.