Artificial intelligence pioneer Anthropic is encountering mounting hurdles in securing sufficient cloud computing capacity from major tech infrastructure providers, as Amazon Web Services (AWS), Google, and Microsoft increasingly view the Claude developer as a direct competitor rather than a traditional customer.
According to details from a confidential S-1 filing recently leaked to The Wall Street Journal, these shifting industry dynamics are complicating Anthropic’s efforts to scale its infrastructure at a time of unprecedented demand for generative AI systems. The leaked document sheds additional light on financial figures previously outlined by Reuters and the Financial Times in late September, revealing the immense capital requirements underpinning the company’s aggressive growth strategy. Notably, the filing indicates that Anthropic projects its cumulative spending on cloud contracts could eventually surpass $750 billion, a figure that includes various major agreements currently undergoing active negotiation.
Despite these astronomical projected outlays, the company continues to grapple with a persistent deficit in available computing power. To bridge the gap, Anthropic has turned to leasing its own specialized data centers and establishing partnerships with third-party data center operators. However, these alternative arrangements have introduced new operational concerns. The filing notes that executive leadership harbors anxieties regarding whether these third-party operators possess a sufficient operating history to guarantee the level of reliability and uptime provided by established cloud hyperscalers.
The company’s immense capital expenditure comes on the heels of a massive financial loss during the first half of the year. Anthropic reported a net loss of $80.7 billion over the initial six-month period, highlighting the staggering costs associated with training cutting-edge frontier models and maintaining massive inference infrastructure. During that same timeframe, the company generated $16.3 billion in revenue. However, financial disclosures within the WSJ report indicate a heavy reliance on a concentrated client base, with just two major customers accounting for roughly one-fifth of total sales.
In addition to infrastructure and profitability challenges, the S-1 filing expands upon previously disclosed catastrophic risk factors. The document explicitly warns that Anthropic could potentially be held legally liable for damages resulting from harmful or unintended AI model behaviors—liabilities that could scale to levels significantly exceeding the company’s available insurance coverage. These risk disclosures underscore the complex legal and operational liabilities facing developers of advanced artificial intelligence systems as their technologies are deployed more widely across commercial and public sectors.
Even as it navigates these profound financial and structural hurdles, Anthropic continues to experience explosive top-line growth. Last week, the company informed prospective investors that its third-quarter revenue alone reached $16.8 billion, representing a robust 45% increase compared to the previous quarter. Buoyed by this strong commercial momentum, Anthropic is reportedly pressing forward with its market debut, targeting a public offering in November.
Representatives for Anthropic did not immediately respond to requests for comment from TechCrunch regarding the leaked S-1 disclosures, the projected cloud infrastructure expenditures, or the timeline for the upcoming public offering.
