Washington's fixation on building the most advanced frontier AI models has left it blind to a more immediate threat: Beijing is quietly capturing the vast majority of the global AI market, particularly in developing nations. While U.S. tech giants compete for bragging rights over the next breakthrough, China is winning customers across Africa, Asia, and Latin America by offering affordable, practical, and easily adaptable AI tools.
Most countries do not need the most powerful model; they need one that works within their budget and infrastructure. China has exploited this gap, pushing its lower-cost, open-weight models like DeepSeek and Alibaba's Qwen into markets that U.S. companies have largely ignored. As a result, Beijing is embedding itself into the digital backbone of the Global South—running hospitals, schools, airports, and government services—and building relationships that will be hard for Washington to dislodge.
The stakes are not merely commercial. When a country's critical infrastructure depends on Chinese technology, it becomes politically vulnerable. Switching suppliers is costly and disruptive, so once Beijing's systems are in place, they tend to stay. This gives China enormous leverage in diplomatic disputes and undermines U.S. efforts to promote its own values and interests. Moreover, Chinese technology often comes with surveillance capabilities and data-security risks, as well as a tendency to echo Beijing's official narratives on history and politics.
China's strategy is deliberate and well-funded. Its Digital Silk Road initiative provides subsidies and financing to help Chinese firms undercut competitors. Huawei, for instance, has expanded its cloud services to nine additional countries in April, offering credits and training in places like Egypt and Brazil. In Egypt, eligible startups received up to $150,000 in cloud credits, along with an Arabic-language model. Such incentives lower the barrier to entry and lock in local developers to Chinese platforms.
The numbers are stark. In Latin America, Chinese data centers now control 25% of Brazil's market, 40% of Chile's, and 100% in Argentina, Mexico, and Peru. Across Southeast Asia, Chinese firms own 57% of Indonesia's data centers and 100% in Thailand, the Philippines, and Malaysia. Microsoft's own data shows DeepSeek usage in Africa was two to four times higher than elsewhere in late 2025, driven by free access and partnerships.
The U.S. response has been anemic. The dismantling of USAID and the lack of a coordinated government strategy have left a vacuum that China is eager to fill. While Washington has started to promote American AI packages and expand export financing, these efforts are too small and too slow to compete with Beijing's deep pockets and aggressive marketing.
To reverse course, the U.S. must treat affordable AI access as a core foreign policy priority. That means significantly increasing funding for developing countries to purchase American technology, investing in local training and developer ecosystems, and supporting tools in local languages. It also means working with partner governments to identify their specific needs and proving that U.S. technology is not only secure but also a better long-term investment.
The race is not over, but time is short. If Washington continues to prioritize frontier models over global adoption, it will wake up to find that Beijing has already set the standards for the rest of the world. That outcome would have profound consequences for American influence, national security, and the future of AI governance. As Orit Frenkel, CEO of the American Leadership Initiative, warns: losing the global AI race would be a strategic defeat with serious repercussions.
