The AI Infrastructure Boom: A Historic Investment Surge
Recent estimates from the Brookings Institution suggest that the current surge in artificial intelligence (AI) is set to become the most significant infrastructure investment in over a century.
According to the Wall Street Journal, economist Stijn van Nieuwerburgh anticipates that investments in data centers and AI-related infrastructure could total around $10.3 trillion from 2025 to 2032, which would account for an average of 3.6 percent of the annual GDP. In a different analysis, Goldman Sachs estimates that AI investment will represent 1.9 percent of GDP in 2026 alone. To put this into perspective, the last time a new industry had a higher economic impact was during the railroad expansion in the late 1800s.
Van Nieuwerburgh’s analysis highlights the scale of this investment. For context, canal construction between 1836 and 1841 reflected an average of 0.66 percent of GDP annually. Railroads from 1870 to 1890 averaged 2.24 percent, electrification from 1905 to 1925 was about 0.5 percent, the interstate highway system from 1956 to 1973 was 1.13 percent, and the telecom and fiber-optic surge from 1996 to 2003 reached 1.1 percent. His projection for AI infrastructure is the highest among these historical benchmarks, at 3.63 percent from 2025 to 2032.
This surge in spending is already apparent. By July, spending on private data center construction reached $37 billion, showing an increase of approximately $9 billion compared to the same period last year, as reported by the Commerce Department. In contrast, overall private construction spending has seen a decline of around $46 billion within the same timeframe.
This shift in investment is beginning to reshape local economies. The Federal Reserve Bank of Richmond noted that construction of data centers is creating challenges in labor availability within its region. In Mississippi, a planned aluminum smelter that promised about 1,000 permanent jobs ultimately went to Oklahoma. A data center near Vicksburg had consumed the electricity that the smelter would have required, according to sources familiar with the decision-making process. Site-selection consultant Didi Caldwell explained, “It’s crowding out manufacturing.”
Some major companies are accumulating a significant portion of this capital expenditure. Analysts predict that tech giants such as Google, Amazon, Meta, Microsoft, and Oracle, the top five hyperscalers, will collectively invest around $4.2 trillion over the four years ending in 2029, according to FactSet. Notably, a substantial amount of this funding is being financed through debt. Van Nieuwerburgh pointed out that many tech companies use off-balance-sheet entities to borrow, which complicates the public understanding of the financial risks involved.
Moreover, skilled trades are also experiencing this impact. In the Washington, D.C. area, the number of unionized electricians has surged from 9,000 to 17,500 in recent years, as noted by Don Slaiman, political coordinator of IBEW Local 26. He remarked that many are coming to the area to pay down college debt. Kwaku Afriyie, 23, who has a degree in cybersecurity, left an entry-level IT job fearing AI would replace him. He retrained as an electrician and now earns about $30 an hour assembling data-center components. Tyler Beam, 28, has spent a considerable time building data centers and recently earned $62 an hour plus overtime, often working 58-hour weeks, which allowed him to buy a GMC Yukon and start searching for a home. “They want to put them up as fast as possible,” he explained.
The emergence of AI data centers is also altering the political landscape. Wynton Hall from Breitbart News has authored a book titled Code Red: The Left, the Right, China, and the Race to Control AI, positioning it as a guide for the MAGA movement to formulate AI policies that benefit society while preventing the entrenchment of leftist power from Silicon Valley or Chinese dominance.
