The United States is currently navigating a rapidly evolving landscape in the global artificial intelligence (AI) race. Despite leading in investment, infrastructure, and advanced AI models, the U.S. position is threatened as Chinese models, including DeepSeek-R1 and the new Kimi K2 Thinking model, demonstrate advantages in cost efficiency and analytical capabilities. With critical implications for sectors such as medicine, cybersecurity, and defense, U.S. policymakers are urged to take decisive steps to maintain the nation’s technological lead.
Chinese authorities are heavily investing in AI development. The government has implemented significant state-backed incentives for companies focused on AI, while also enhancing energy subsidies for data centers. Notably, a $37 billion investment is planned for a large-scale data center in Wuhu, which is part of a broader push outlined in China’s 15th Five-Year Plan. According to Goldman Sachs, Chinese cloud service providers are projected to increase capital expenditures by 65 percent in 2025, with a total of $70 billion allocated for further development.
The competitive landscape has shifted, with Moonshot AI’s Kimi K2 Thinking model surpassing OpenAI’s GPT-5 and Anthropic’s Claude Sonnet 4.5 in performance tests. This achievement highlights a concerning trend for American AI developers, as these Chinese models are less costly to produce than their U.S. counterparts. The implications extend beyond commercial competition; China aims to embed AI systems at the core of its national security strategies, raising significant concerns for U.S. interests.
Research from Shanghai Jiao Tong University underscores this ambition, as the institution secured a government contract to explore the use of AI in automated military strategies, such as “kill webs” that adapt dynamically in combat scenarios. In another example, AI surveillance tools in Xinjiang are employed to monitor citizens, leveraging behavioral and personal data to assess suspicion levels, often resulting in reports triggered by minor actions like using a VPN.
To counteract these developments, the United States must foster an environment conducive to innovation in AI. This includes reducing regulatory burdens and providing robust government support. There are signs of progress, illustrated by OpenAI’s plans to construct a $500 billion data center in Texas, which significantly outstrips investments seen in China.
Congress plays a critical role in this equation. It is crucial that lawmakers consider reinstating a proposed ten-year ban on state-level AI regulations, which was omitted from the Tax Cut and Spending Bill in July 2023. Over-regulation can stifle growth and inadvertently benefit international competitors like China, which is vying for dominance in AI technology.
The United States is home to the world’s most advanced AI chip, the Nvidia Blackwell. By restricting access to this technology, the U.S. can hinder China’s potential to close the gap in AI innovation. DeepSeek’s recent challenges in accessing Nvidia’s chips serve as a case in point, as delays in model releases highlight the impact of technological limitations.
Ultimately, the AI race is not merely about technological superiority; it encapsulates the future of global power across multiple domains, including defense and healthcare. As China increasingly weaponizes AI, setting concerning precedents, the United States must remain vigilant and proactive. To sustain its leadership, U.S. policymakers must continue to prioritize innovation while ensuring that competitors do not gain access to essential technologies.
Samuel Field is a policy fellow at the Pinsker Centre, a British-based foreign policy think tank focused on the Middle East and wider international affairs.
