Construction spending on AI data center shells surged in July, increasing at an annualized rate of almost 60% compared to July of the previous year. This data from the U.S. Census Bureau indicates a growing frenzy in AI infrastructure development during the summer months. Spending on data center construction reached an annual pace exceeding $75 billion in July. This figure specifically captures the costs associated with building the large, warehouse-like structures designed to house AI technology. These "shells" are distinct from the internal components that power AI functionality. The expense of filling these data centers with servers, processors, and memory chips is significantly higher than their construction costs. Estimates suggest that construction accounts for only about 20% of the total investment required for an AI data center. The substantial spending highlights the rapid expansion of AI capabilities. This growth is occurring despite increasing political scrutiny and potential backlash. The robust construction spending reflects a strong underlying demand for AI infrastructure.