In the high-stakes race to power the artificial intelligence boom, China’s infrastructure advantages are leaving the U.S. scrambling, according to experts who recently toured the country’s vast energy networks. Longtime Chinese energy analyst David Fishman, speaking to Fortune, likened China’s setup to hitting “grand slams” while the U.S. can only “get on base.” This disparity stems from China’s aggressive buildout of renewable energy and grid capacity, tailored to feed the insatiable hunger of data centers driving AI advancements.

Fishman’s observations came during a trip organized by the Asia Society Policy Institute, where U.S. delegates witnessed firsthand how China has integrated massive solar and wind farms with ultra-high-voltage transmission lines. These lines span thousands of miles, efficiently delivering power to data center hubs in remote areas like Inner Mongolia, where electricity is abundant and cheap. In contrast, the U.S. grid, fragmented by regional operators and regulatory hurdles, struggles to keep pace with surging demand from tech giants like Microsoft and Google.

The Grid Gap Widens

Recent projections underscore this chasm. Data centers in the U.S. could consume up to 12% of national electricity by 2030, as noted in a report from WebProNews, while China’s renewable integration allows it to handle similar loads with less strain. X posts from industry watchers, such as those by financial analyst Shanu Mathew, highlight how U.S. utilities face a 7-26% generation increase by 2028, with 44% of that load growth attributed to data centers—demanding over $2 trillion in global energy investments.

China’s edge isn’t just in scale; it’s in speed. The country added 277 gigawatts of power capacity recently, dwarfing U.S. efforts, according to posts on X by host Mario Nawfal. This has enabled China to build hundreds of AI-focused data centers, though some now sit underutilized due to overcapacity, as detailed in a MIT Technology Review analysis. Yet, this surplus positions China to scale AI training rapidly, unhindered by blackouts or delays that plague U.S. operators.

Regulatory Roadblocks and Innovations

U.S. grid bottlenecks are exacerbating the issue. Regional operators have delayed upgrades mandated by the Federal Energy Regulatory Commission in 2021, aimed at dynamically updating transmission capacities based on real-time weather, per a report from Data Center Dynamics. Incidents like last summer’s mass switch to on-site generators in Virginia’s Data Center Alley, which consumes as much power as Boston, alarmed utilities, as reported by Reuters.

Meanwhile, China is pushing forward with plans like Shanghai’s target of five new data centers in 2025 to boost AI computing, reaching 246 exaflops nationally—second only to the U.S., according to the South China Morning Post. This growth, however, raises emissions concerns, with AI-driven data centers surging energy demand, as outlined in a Carbon Brief briefing.

Implications for Global Tech Dominance

The power crunch is reshaping corporate strategies. Big Tech firms are eyeing nuclear restarts and microgrids in the U.S., but experts warn this patchwork approach can’t match China’s centralized planning. A ZeroHedge post on X estimates a 36-gigawatt shortfall in U.S. data center power by 2028, forcing delays in AI deployments.

For industry insiders, the lesson is clear: without bold infrastructure reforms, the U.S. risks ceding AI leadership. China’s model, blending state-driven investments with green tech, offers a blueprint—but adapting it to America’s decentralized system will require unprecedented collaboration between regulators, utilities, and tech leaders to bridge the gap before it’s too late.