China is rapidly advancing in artificial intelligence (AI), yet a significant disparity remains in AI infrastructure spending between the United States and China. By the end of 2027, US hyperscalers are projected to invest approximately 8.3 times more than their Chinese counterparts in AI infrastructure, including critical resources such as GPUs and data centers. This substantial financial gap enables US companies to train more sophisticated models, attract top talent, and deploy AI systems on a much larger scale compared to Chinese firms, thereby enhancing their overall competitive advantage in the AI landscape.

Callum: Callum Williams is the San Francisco bureau chief and senior economics writer at The Economist, with a focus on AI, financial markets, real estate, and economic history. He recently drew attention to the wide gap in AI infrastructure spending between the US and China, noting the compute advantages that favor American hyperscalers. His analysis frames the structural factors enabling greater scale in US AI model training, inference, and developer ecosystems.

AI Infrastructure: US companies hold a clear advantage in securing GPUs, data centers, power, networking, and related resources needed for large-scale AI systems.
Global Competition: Differences in capital deployment create broader opportunities for US AI firms to train advanced models, attract talent, and deliver products at scale relative to Chinese competitors.
Technological Scaling: While algorithms contribute to progress, infrastructure capacity determines the ability to experiment extensively and subsidize widespread AI adoption.