The competition between the United States and China in the field of artificial intelligence (AI) has entered a heated stage. Represented by DeepSeek, AI’s extensive use in China, including in the military sector, is threatening the national security of the United States. U.S. Director of National Intelligence, Tulsi Gabbard, warned that if the Chinese Communist Party (CCP) achieves its goal of AI dominance by 2030, the U.S. may face comprehensive passive measures ranging from economic to military aspects.
With the emergence of DeepSeek, the CCP has issued a call for “AI for all.” Tech giants like Baidu, Alibaba, and Tencent have launched no less than 10 low-cost models in the past two weeks, covering areas such as computer vision, robotics, and image generation, forming a collective offensive to weaken the market advantage of high-end U.S. products. China is attempting to seize global AI standard-setting authority and market share through open-source AI models, prompting the Trump administration to take countermeasures.
Balaji Srinivasan, a former partner at Andreessen Horowitz, recently stated that the CCP is adopting its usual strategy with AI: learn, imitate, optimize, and then bankrupt all competitors by offering low prices and massive scale.
Meanwhile, Chinese cloud service providers that carry AI development work are significantly reducing prices, and this competition is spreading to the global market. Kevin Xu, founder of Interconnected Capital and tech investor, believes that this is a natural evolution of China’s ecosystem in price wars extending to other markets.
With breakthroughs in generative AI models, artificial intelligence is rapidly infiltrating various corners of the military domain, with intelligent warfare becoming a significant trend in future conflicts. A report by the South China Morning Post on March 23 revealed China’s ambitions to integrate AI into its military system.
According to Sam Bresnick, a researcher at Georgetown University’s Center for Security and Emerging Technology (CSET), the Chinese military is exploring the potential value of AI in military decision-making, and the emergence of advanced models like DeepSeek R1 may aid China’s development in this area.
In recent years, the Chinese military has been advocating for the integration of high-end technologies, especially AI, into its military capabilities to enhance combat effectiveness. This includes improving the efficiency of drone swarm tactics, pilot training effectiveness and realism, and battlefield decision support.
Military analysts believe DeepSeek will accelerate the CCP’s military intelligence process. DeepSeek may have already ventured into operational areas such as intelligence reconnaissance, battlefield decision-making, and drone swarm tactics.
Gabbard warned that the CCP is attempting to leverage AI technology and target its assets in the space domain. Additionally, the Chinese military may use AI to create fake news or launch cyber attacks.
In a report released on March 25, Gabbard stated that the CCP plans to surpass the U.S. in becoming the top AI power by 2030, while China’s military is deploying hypersonic weapons, stealth aircraft, and cyber assets capable of striking the U.S. mainland.
The launch of DeepSeek and its extensive use in China coincides with the intense technological competition between the U.S. and China in the AI field, which has evolved beyond a mere battle of technology and markets to a strategic game concerning national security and global leadership.
Currently, the U.S. maintains an advantage in the field of AI, including algorithms and hardware products. Although China’s tech threat is significant, it has not yet shaken the fundamental advantage of the U.S.
The U.S.’s accumulation in basic research and high-end hardware design remains a barrier that China finds challenging to overcome. NVIDIA’s GPUs, Google’s TPUs, and xAI’s computing framework represent the top levels of AI hardware, while China still lags behind in chip manufacturing processes compared to TSMC and U.S. domestic manufacturers. AI training and inference rely on high-performance GPUs like the NVIDIA H100, A100, and the U.S. has significantly weakened China’s competitiveness in AI training by restricting chip exports. Even though China has companies like Huawei and Cambricon attempting to develop AI chips, they still lag behind NVIDIA and AMD in process technology, software ecosystem, and overall performance.
Moreover, the leading position of U.S. universities and research institutions (such as MIT, Stanford) in AI algorithm and theory innovation is a challenge for China to surpass in the short term. DeepSeek’s success is more about engineering optimization rather than fundamental breakthroughs. While its open-source model is competitive, the U.S. still holds the initiative in the cutting-edge closed-source field (such as OpenAI’s o3 or xAI’s next-generation models).
However, while OpenAI is considering open-sourcing some technologies to counter China’s competition, they are also concerned about potentially losing their high-priced model in the high-end market.
The focus now is on whether the U.S.’s decentralized response mechanism and high-cost commercial model can withstand China’s efficient onslaught. Gabbard’s warning is not baseless: if China achieves its goal of AI dominance by 2030, the U.S. may face comprehensive passive measures from economic to military aspects.
The U.S.-China competition has also expanded into the industrial robots domain, with the U.S.’s previous leading position no longer existing, as none of the four major industrial robot arm suppliers are located in the U.S. According to the latest data from the International Federation of Robotics (IFR), China currently has approximately 1.8 million industrial robots in use, making it the world’s largest industrial robot market.
Facing China’s rapid development of AI-dominated industrial robots, several U.S. companies have called on the U.S. government to quickly establish a national robot strategy and set up a federal-level agency to address China’s challenge.
On March 26, a group of robot companies including Tesla, Boston Dynamics, Agility Robotics, Apptronik, and others met with members of Congress on Capitol Hill to showcase their latest research outcomes and discuss the necessity of establishing a federal office amid the rapid development of AI and strategic policy against global competitors (China).
The Association of Advanced Automation, based in Ann Arbor, Michigan, agrees on the necessity of formulating a national strategy and points out that China and some other countries have already implemented such policies. In a statement, the association warned that without a comprehensive plan, the U.S. could not only lose the robot race but also the AI race.