The US-China AI blocs taking shape around advanced chips, data centers, software and technical standards are creating a difficult choice for countries that want access to both technology ecosystems. Washington is building a network of trusted partners, while Beijing is promoting a broader international cooperation platform aimed especially at developing countries.
That does not mean the world has split neatly into two camps. Many governments are trying to cooperate with both powers at once, using American technology for the most sensitive systems while considering Chinese models, equipment or investment where cost and availability matter more. But the room for that balancing strategy may be shrinking.
What the US-China AI blocs are building
The U.S. approach centers on controlling the most strategically important parts of the AI supply chain. Advanced processors, semiconductor manufacturing equipment, cloud computing capacity and the infrastructure needed to run large models have become closely tied to national security policy.
In May 2025, the U.S. Department of Commerce said it was rescinding the Biden-era AI Diffusion Rule while strengthening other controls on advanced semiconductors. The department said the United States would pursue a strategy of working with trusted foreign countries while restricting access by adversaries. The policy illustrates the basic tension for Washington: expand the global reach of American AI while limiting the transfer of technology that could strengthen China’s military or industrial capabilities.
The U.S. strategy has since been linked to broader efforts to secure supply chains. The White House describes Pax Silica, launched in December 2025, as an effort by the United States and trusted partners to develop resilient semiconductor and critical-mineral networks that support the AI economy.
China is presenting a different model. At the 2025 World Artificial Intelligence Conference in Shanghai, Chinese officials proposed the World Artificial Intelligence Cooperation Organization, or WAICO. China says the organization would support innovation, technical cooperation, AI capacity building and broader participation by countries in the Global South.
In an official statement, China’s Foreign Ministry said WAICO would complement United Nations efforts and promote cooperation based on sovereignty, equality and what it calls an open and inclusive approach. The initiative also fits Beijing’s wider push for global AI governance rules that emphasize access, development and national control.
Why middle powers are reluctant to pick one side
For many countries, the decision is less ideological than practical. Building an AI economy requires far more than choosing a chatbot. Governments need reliable electricity, data centers, high-speed networks, specialized chips, engineering talent, financing and rules for handling sensitive data.
The United States remains a critical source of advanced processors, cloud services, research partnerships and leading AI companies. It also has longstanding security relationships with countries across Europe, Asia and the Middle East. China, meanwhile, can offer lower-cost hardware, open-source models, infrastructure financing and commercial relationships that may be attractive to governments trying to expand digital services quickly.
That helps explain why some countries are pursuing what officials describe as a flexible or multi-aligned approach. Indonesia, for example, has sought benefits from both major powers while joining China’s proposed AI cooperation organization, according to the Associated Press. Singapore has deep technology and security links with the United States but has also considered engagement with China’s platform.
Saudi Arabia illustrates another version of the strategy. The kingdom has explored Chinese AI technology while also seeking access to advanced American chips and computing infrastructure. Japan is closely aligned with Washington on security and semiconductor policy, but officials and industry participants still recognize that Chinese AI systems may be useful for applications that are not considered high risk.
France is another important case because it is trying to preserve a European path through its own AI companies and regulatory choices. The country’s position shows that the emerging contest is not necessarily a binary U.S.-China system. Some governments want a third option that gives them greater control over data, standards and industrial policy.
The pressure is moving from diplomacy to infrastructure
The most consequential choices may happen quietly, through procurement contracts and technical standards. A government that builds its data centers around one supplier’s chips, networking equipment and software tools may find it expensive to switch later.
That creates a form of technological path dependence. Once a country trains workers on a particular platform, stores data in a specific cloud environment and adopts a certain safety or certification system, changing direction can take years. The decision may affect public-sector software, telecommunications networks, universities, banks and defense contractors at the same time.
Washington is likely to ask partners for stronger safeguards against diversion of advanced chips and technology to China. Beijing is likely to emphasize that U.S. restrictions create an unequal global technology system and that developing countries should not have to accept rules written by major powers.
Both messages have an audience. Countries want secure access to frontier technology, but they also want affordable tools and freedom to set their own development policies. If the United States offers security but too little access, partners may look elsewhere. If China offers access but raises concerns about transparency, data control or political influence, governments may limit cooperation to selected areas.
What to watch next
The next phase of the rivalry will be measured by concrete commitments rather than speeches. Watch for which countries sign formal agreements tied to advanced chips, data centers and critical minerals. New export-control rules will show how far Washington is willing to extend the definition of a trusted partner. The development of WAICO will reveal whether China can turn its proposal into a durable institution with meaningful participation.
Another important test will be whether countries can separate high-risk and low-risk uses of AI. A government might rely on U.S. suppliers for military, intelligence or critical-infrastructure systems while using Chinese or locally developed models for translation, customer service, education or other applications. That kind of selective alignment could become the dominant pattern.
The result may not be a clean division into two global AI blocs. Instead, the world could develop overlapping technology networks, with countries switching partners by sector and risk level. But as chips, software and infrastructure become more tightly connected to national security, keeping one foot in each camp will require increasingly careful diplomacy.
For countries caught between Washington and Beijing, the central question is no longer simply which AI is better. It is how to gain the benefits of both systems without becoming dependent on either one.
Leave a Reply