China is rapidly outpacing the United States in the race to secure energy resources crucial for the development of artificial intelligence, according to the leading AI company Anthropic. The San Francisco-based firm has warned that without urgent reform to speed up infrastructure projects, the US risks falling further behind in its technological rivalry with China.
In a report released on 22 July, Anthropic highlighted the growing gap in power generation capacity between the two countries, citing data from Australian think tank Climate Energy Finance. In 2023, China added 400 gigawatts of new power capacity, while the US said only a fraction of that—just a few dozen gigawatts, or roughly one-tenth of China’s total.
The report emphasised the looming challenges for the US AI sector, which is projected to need at least 50 gigawatts of power capacity by 2028. This discrepancy in energy infrastructure, Anthropic said, poses a serious concern for the future of AI innovation in the United States.
Energy infrastructure becomes a key battleground
Energy availability is emerging as a pivotal issue in the escalating AI competition between the world’s two largest economies. While the US continues to invest heavily in AI hardware such as semiconductors, China is directing a significant share of its AI spending towards building the data centres and energy infrastructure needed to support its ambitions.
“China – also vying for AI leadership – does not face the same set of regulatory constraints that we do,” Anthropic noted in the report. Although both nations require permits for large-scale infrastructure projects, Chinese regulators process applications at a much faster pace, the company said.
Matty Zhao, co-head of China equity research at Bank of America Securities said in an interview that China is prioritising long-term support for AI growth through energy investments.
This emphasis on energy development is already taking shape in major projects. Last week, China announced a new mega dam in Tibet, estimated to cost 1.2 trillion yuan (US$167 billion), which will become the world’s largest hydropower facility. The dam is expected to generate 300,000 gigawatt-hours of electricity annually.
The US faces mounting concerns over energy readiness
Leading figures in the US technology sector have voiced concern about the nation’s ability to keep pace with energy demands driven by AI advancements. In February, Google’s global head of energy markets development and innovation, Caroline Golin, described the situation as a “capacity crisis”.
Elon Musk, founder of Tesla and AI start-up xAI, also issued a warning in May, stating that the US could encounter serious energy supply shortages for AI applications by mid-2026. His comments, reported by CNBC, added to growing alarm among tech executives.
According to a separate analysis by UK-based energy think tank Ember, China is leading global efforts in renewable energy. In 2024 alone, the country was responsible for 71 per cent of the worldwide increase in hydropower generation. It also added more solar and wind capacity than the rest of the world combined.
In contrast, the US has been hampered by bureaucratic hurdles, particularly in building essential transmission lines and interconnection infrastructure. Anthropic emphasised that these regulatory challenges must be addressed urgently to prevent further delays in scaling AI systems.
“The US is not on track to meet the energy needs of AI training or inference by 2028,” the report concluded. The company urged policymakers to streamline existing regulations and expedite project approvals to prevent long-term damage to the nation’s competitive position in artificial intelligence.
