Anthropic has identified its reliance on Amazon and Google as a significant business risk. These companies act simultaneously as infrastructure providers, key customers, and direct competitors in the AI market. In 2025, up to 47% of the startup's services were sold via Amazon Web Services and Google Cloud.
The financial data reveals a rapid increase in revenue concentration. In 2023, Amazon and Google accounted for 11% of Anthropic's revenue, rising to 32% in 2024. Furthermore, the company's two largest clients each contributed 12% of total revenue last year, highlighting a high level of dependency on a few major partners.
Anthropic reported a 12-fold revenue increase to $4,6 billion last year, with $3,8 billion coming from Claude services and $789 million from subscriptions. Cloud marketplaces generated 47% of total revenue, or $2,16 billion, while the company paid $351 million in platform commissions.
The startup is heavily investing in infrastructure. By early 2026, its long-term commitments to data center providers reached $417 billion, securing 3,5 GW of computing power. Anthropic notes that while partnerships with Amazon, Google, and Microsoft accelerate growth, they also grant these firms access to sensitive pricing and marketing data.
Beyond infrastructure risks, Anthropic warns of legal uncertainties regarding autonomous AI agents. The company notes that current regulations do not define the status or liability of these agents, increasing the potential for security breaches, errors, and legal claims that existing contract liability limits may not cover.
