US-China AI Co-opetition* and Taiwan’s Role
At the September summit between Donald Trump and Xi Jinping, artificial intelligence emerged as one of the key topics of discussion between the United States and China. More recently, President Trump invited several of the top U.S. artificial intelligence executives to discuss the field’s prospects, underscoring AI’s extraordinary importance.
The U.S.-China AI rivalry has been expanding from a contest over model performance to a broader struggle encompassing computing power, capital, talent, industrial applications and international norms. The U.S. is strong in frontier R&D and capital mobilization, while China has become more competitive by narrowing the gap in AI models, expanding applications and open ecosystems. The two sides are competing for technological leadership, but they also face shared risks. A key question in this rivalry is how Taiwan can draw on its own strengths to become a vital hub connecting innovation with production.
The AI technology gap between the U.S. and China is closing, with each country having its own strengths. Stanford University’s 2026 AI Index Report notes that American and Chinese models have repeatedly traded the lead since early 2025, with the U.S. continuing to release more top-tier models and China dominating in terms of research papers and patent output. Rankings can reflect only certain aspects of AI capabilities; the real competition also includes system reliability, service costs and commercial returns.
The U.S. advantage lies in the mutually supportive dynamic between research, investment and infrastructure. Abundant capital allows for long-term R&D and large-scale training but also creates pressure to deliver returns on investment. On the other hand, China’s key opportunity lies in converting model capabilities into tangible benefits for industries such as manufacturing, logistics and retail. However, an abundance of use cases does not automatically translate into productivity growth; improvements in data quality, business processes and talent allocation are still necessary. Ultimately, both countries need to answer the same question: Can AI continue to create more value than the investment it requires?
Chips and computing power remain significant constraints in the competition. U.S. restrictions on advanced chip exports have made it more difficult for China to acquire high-end computing power, but they have also driven the development of domestic alternatives in China. However, the competition over computing power involves chips, manufacturing, software tools and cluster efficiency, and no breakthrough in any one area is sufficient to solve all the others. The U.S. can increase its rival’s costs through restrictions, while China, for its part, can alleviate the pressure of those restrictions through algorithmic optimization and engineering improvements. The two sides’ relative advantages will continue to shift.
At the same time, competition has not eliminated the need for cooperation: Deepfakes, fraud and the loss of control over automated systems can all carry cross-border repercussions. Even if both sides remain guarded about their core technologies, there is still reason to establish communication around incident reporting, risk assessment and misuse prevention. Whether relations are improving should be judged by whether these specific mechanisms can ultimately be put into practice.
In this context, Taiwan can transform its hardware-industry strengths into a stabilizing role in the supply chain by emphasizing the importance of cross-strait industrial complementarity and peace across the Taiwan Strait. In this way, it can make its capabilities a necessary condition for the continued operation of international industries. To bring this about, Taiwan needs to further extend its manufacturing strengths into systems integration and application services, for example by combining chips, servers and industrial software to provide businesses with solutions that deliver tangible benefits. Only in this way can Taiwan increase added value: by channeling more of its industrial earnings into local R&D, talent development and the upgrading of small and medium-sized enterprises, thereby reducing its dependence on the fortunes of any single product.
The future of U.S.-China AI relations is more likely to be characterized by a long-term coexistence of competition and limited cooperation. If Taiwan is to capitalize on the opportunities ahead, it must continue to strengthen its technological capabilities, supply reliability and institutional trust. A truly secure role as a hub comes from the ability to consistently address the problems the industry faces while demonstrating to international partners that maintaining cooperation and stability across the Taiwan Strait is in everyone’s interest.
*Editor's note: Co-opetition refers to a situation in which people or businesses or governments might engage in both cooperation and competition at the same time, working toward a common goal.
The U.S.-China AI rivalry has been expanding from a contest over model performance to a broader struggle encompassing computing power, capital, talent, industrial applications and international norms. The U.S. is strong in frontier R&D and capital mobilization, while China has become more competitive by narrowing the gap in AI models, expanding applications and open ecosystems. The two sides are competing for technological leadership, but they also face shared risks. A key question in this rivalry is how Taiwan can draw on its own strengths to become a vital hub connecting innovation with production.
The AI technology gap between the U.S. and China is closing, with each country having its own strengths. Stanford University’s 2026 AI Index Report notes that American and Chinese models have repeatedly traded the lead since early 2025, with the U.S. continuing to release more top-tier models and China dominating in terms of research papers and patent output. Rankings can reflect only certain aspects of AI capabilities; the real competition also includes system reliability, service costs and commercial returns.
The U.S. advantage lies in the mutually supportive dynamic between research, investment and infrastructure. Abundant capital allows for long-term R&D and large-scale training but also creates pressure to deliver returns on investment. On the other hand, China’s key opportunity lies in converting model capabilities into tangible benefits for industries such as manufacturing, logistics and retail. However, an abundance of use cases does not automatically translate into productivity growth; improvements in data quality, business processes and talent allocation are still necessary. Ultimately, both countries need to answer the same question: Can AI continue to create more value than the investment it requires?
Chips and computing power remain significant constraints in the competition. U.S. restrictions on advanced chip exports have made it more difficult for China to acquire high-end computing power, but they have also driven the development of domestic alternatives in China. However, the competition over computing power involves chips, manufacturing, software tools and cluster efficiency, and no breakthrough in any one area is sufficient to solve all the others. The U.S. can increase its rival’s costs through restrictions, while China, for its part, can alleviate the pressure of those restrictions through algorithmic optimization and engineering improvements. The two sides’ relative advantages will continue to shift.
At the same time, competition has not eliminated the need for cooperation: Deepfakes, fraud and the loss of control over automated systems can all carry cross-border repercussions. Even if both sides remain guarded about their core technologies, there is still reason to establish communication around incident reporting, risk assessment and misuse prevention. Whether relations are improving should be judged by whether these specific mechanisms can ultimately be put into practice.
In this context, Taiwan can transform its hardware-industry strengths into a stabilizing role in the supply chain by emphasizing the importance of cross-strait industrial complementarity and peace across the Taiwan Strait. In this way, it can make its capabilities a necessary condition for the continued operation of international industries. To bring this about, Taiwan needs to further extend its manufacturing strengths into systems integration and application services, for example by combining chips, servers and industrial software to provide businesses with solutions that deliver tangible benefits. Only in this way can Taiwan increase added value: by channeling more of its industrial earnings into local R&D, talent development and the upgrading of small and medium-sized enterprises, thereby reducing its dependence on the fortunes of any single product.
The future of U.S.-China AI relations is more likely to be characterized by a long-term coexistence of competition and limited cooperation. If Taiwan is to capitalize on the opportunities ahead, it must continue to strengthen its technological capabilities, supply reliability and institutional trust. A truly secure role as a hub comes from the ability to consistently address the problems the industry faces while demonstrating to international partners that maintaining cooperation and stability across the Taiwan Strait is in everyone’s interest.
*Editor's note: Co-opetition refers to a situation in which people or businesses or governments might engage in both cooperation and competition at the same time, working toward a common goal.
Previous article![]( "Trump Is Yielding a Key Country to Iran*")
