Anthropic scales AI compute to meet rising global demand
Expansion of compute infrastructure supports rising demand for Claude models, improving scalability and serving a growing global customer base.
AI company Anthropic has announced a major expansion of its compute infrastructure through a new partnership with Google and Broadcom, securing multiple gigawatts of next-generation TPU capacity expected to come online from 2027.
The increased compute supply is intended to support its frontier Claude models and meet rapidly growing global demand.
The company said the expansion reflects a continued strategy of scaling infrastructure to match accelerating customer growth. Demand for Claude has increased sharply in 2026, with revenue run-rate surpassing $30 billion and the number of high-spending business customers doubling in a short period.
Most new computing capacity will be based in the United States, aligning with broader investment plans in domestic AI infrastructure. The partnership builds on collaborations with Google Cloud and Broadcom, alongside continued use of multiple hardware platforms to improve performance and resilience.
Anthropic stated that diversifying compute across different providers helps optimise workloads and maintain reliability for enterprise users. Claude remains available across major cloud platforms, supporting its position in a competitive and rapidly scaling AI market.
The expansion reflects how rapidly growing demand for advanced AI systems is driving large-scale investment in underlying compute infrastructure, with potential implications for capacity, reliability, and the global distribution of AI development resources over time.
It also suggests how access to computing resources is becoming a key factor shaping competitiveness and innovation across the AI ecosystem.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!