The AI race is no longer just about algorithms and chips, but about who controls the energy that feeds it. China has a huge advantage, which it has established with huge investments in hydropower, nuclear power, and renewable energy sources. Thus, it has access to abundant and cheap electricity and sometimes boasts 80-100% grid reserves. AI’s data centers are not seen as a threat to grid stability, but as a way to absorb electricity oversupply.
In contrast, the US network is in serious trouble due to the surge in demand for artificial intelligence. Rising costs, bottlenecks, and patchy expansion are forcing private companies to build their own microgrids, seek nuclear agreements, or even run gas generators, all just to keep AI servers running.
China benefits from technocratic, long-term planning. Energy generation, transmission and AI infrastructure are coordinated by centralized five-year plans, ensuring that energy systems anticipate demand instead of catching up. The United States, meanwhile, suffers from fragmented regulation, lengthy permitting procedures and a capital market that prioritizes quick returns, making it difficult to build the robust, future-proof energy grids required by AI.
As a result, at present, America cannot compete effectively on the energy infrastructure front. In the United States, the growth of artificial intelligence is increasingly tied to debates about data center energy consumption and grid limitations, which is in stark contrast to China.
Translated and edited by Alex Kada