Global AI race: China leads in R&D spending as US infrastructure investment surges

Global technological competition is intensifying as investment shifts toward physical AI infrastructure and research and development. According to the United Nations' Global Innovation Index 2026, China has surpassed the United States in total R&D spending, investing approximately $860 billion in purchasing power parity compared to $850 billion from the US. China also accounts for 37% of global scientific publications, while the US holds 16%.

In the United States, AI infrastructure investment is reaching unprecedented levels. Research from Columbia Business School indicates that between 2025 and 2032, US investment in data centers, power facilities, and AI hardware could reach $10.3 trillion, averaging 3.63% of US GDP annually. This scale exceeds historical infrastructure expansions, such as the US railroad construction of the late 19th century.

This massive spending presents financial challenges for major US tech firms. For companies including Microsoft, Alphabet, Amazon, Meta, and Oracle, capital expenditures are projected to exceed operating cash flow by 2026. Furthermore, these firms are increasingly relying on external financing and debt to fund high-cost projects, such as Meta’s Hyperion AI data center. The technological race is increasingly defined by the ability to scale physical assets like semiconductors, data centers, and electricity networks.