From Power to Chips: Trump Taps Allied Money for AI Race With China

The U.S. strategy, as reported, appears to use allied investment to bolster its domestic power infrastructure, particularly in regions with high AI data center density. South Korea's commitments include a $23.3 billion gas-fired combined-cycle power plant in Encinal, Texas, with commercial operation planned in phases by 2032. Additionally, a "Korea-U.S. nuclear framework" involves the construction of eight large nuclear reactors in the United States, with at least two using the Korean-designed APR1400, supported by $120 billion from a U.S. investment fund.

Japan's agreements also prioritize power, including a $17 billion natural gas plant in Pennsylvania, a region where AI power demand is noted to be rising sharply and where Amazon plans a large AI and cloud data center campus. Small modular reactors led by a U.S.-Japan joint venture, GE Vernova Hitachi, are also planned for Tennessee and Alabama, states that already host major data centers operated by companies such as SpaceX, Google, and Meta.

This focus on power generation highlights a reported challenge for the United States in its AI competition with China. While the U.S. is said to lead in AI accelerators and software, China reportedly generates more than double the electricity. A Citadel Securities report suggests China will add nearly six times more generation capacity than the U.S. over the next five years, while the U.S. may contend with grid interconnection and permitting issues, potentially affecting its AI development pace.