A leaked S-1 filing reveals Anthropic cannot secure sufficient cloud capacity from AWS, Google, and Microsoft because these hyperscalers increasingly view the Claude-maker as a competitor rather than a customer. The document, obtained by The Wall Street Journal, indicates the company estimates it could spend more than $750 billion on cloud contracts, yet this demand remains unmet. Anthropic is now leasing its own data centres and partnering with third-party operators, though the firm worries these providers lack the operating history required for reliable service. This infrastructure struggle follows an $80.7 billion net loss in the first half of the year, despite $16.3 billion in revenue where just two customers accounted for roughly a fifth of sales. The filing also flags potential liability for damages from harmful AI behaviour that could exceed available insurance coverage. While the company recently reported third-quarter revenue of $16.8 billion, a 45% rise from the previous quarter, it is reportedly targeting a November IPO despite these operational headwinds.
The situation matters because it exposes a structural tension between an AI builder and the very infrastructure providers it depends on. As Anthropic scales, it risks facing supply constraints that could stall its growth plans or force reliance on less proven data centre partners. The financial exposure from potential AI liability further complicates the path to a public listing, creating uncertainty for investors regarding long-term stability.
- Two customers currently account for about a fifth of total sales.
- Third-quarter revenue reached $16.8 billion, up 45% from the previous quarter.
- The company is targeting an initial public offering in November.
