China’s AI models gain global ground, challenging US dominance in artificial intelligence

Alibaba’s Qwen ranks among the top 15 models globally. (Courtesy: Reuters photo)

Chinese artificial intelligence companies are quickly gaining traction in global markets, loosening America’s hold on the industry and accelerating what some analysts now call a technological Cold War. From Europe to Africa to Southeast Asia, institutions are beginning to test and adopt Chinese-developed large language models such as DeepSeek and Alibaba’s Qwen—often choosing them over US products like ChatGPT because of lower cost, open-source flexibility, and expanding local access, the Wall Street Journal reported.

In a significant shift, global banks such as HSBC and Standard Chartered are testing DeepSeek models, while Saudi Aramco has installed the software in its main data center. Even US tech giants including Amazon, Google, and Microsoft are offering DeepSeek to customers via their platforms—even though the White House has banned its app on some government devices due to security concerns.

Chinese AI is nearly as powerful—and far cheaper

While OpenAI’s ChatGPT remains the dominant AI model worldwide—with 910 million global downloads compared to DeepSeek’s 125 million—China’s products are closing the gap in both performance and scale. DeepSeek’s newest model ranks third globally in performance, tied with Elon Musk’s Grok 3, according to recent benchmarks. Alibaba’s Qwen and Tencent’s Hunyuan models also rank among the top 15 globally.

China’s edge lies in two foundational areas: a vast pool of human capital and huge amounts of training data. These factors have helped its models evolve rapidly and made it easier for foreign users to access high-quality tools at a fraction of the price. For example, businesses on the platform Latenode say DeepSeek performs “overall the same” as top US models but is 17 times cheaper.

From Tokyo to Johannesburg, global adoption rises

In Tokyo, startup Abeja selected Alibaba’s Qwen to develop custom AI models for Japan’s Ministry of Economy. In South Africa, researchers at the University of the Witwatersrand chose DeepSeek because it works offline and doesn’t risk exposing sensitive university data—an important edge for institutions focused on privacy.

These examples reveal a larger trend: the rise of “AI alignment,” where countries and companies may increasingly need to pick between American and Chinese ecosystems. “Whoever gets there first will be difficult to supplant,” Microsoft President Brad Smith warned during a Senate hearing.

The threat of fragmentation—and manipulation

This global split isn’t just economic—it has geopolitical and ethical consequences. US lawmakers are sounding the alarm over the risk that Chinese models could spread government-approved misinformation and censorship. For instance, the consumer version of DeepSeek has been found to avoid sensitive topics like Tibet and Xinjiang, aligning with Communist Party narratives.

Industry experts say the fracture could reduce America’s ability to monitor and shape AI development globally. “If they are dependent on the global ecosystem, then we can govern it,” said Ritwik Gupta, AI policy fellow at UC Berkeley. “If not, China is going to do what it is going to do, and we won’t have visibility.”

US AI faces global pushback—and opportunity loss

As more governments and companies worldwide adopt Chinese models, American firms risk losing both revenue and influence. For example, when the US blocked exports of Nvidia’s H20 chip to China, analysts estimated a $10 billion loss in revenue. The longer-term concern is that Beijing could use its own AI platforms to shape digital infrastructure and standards in developing countries—securing future influence and partnerships that once defaulted to the West.

OpenAI, Google, and Meta are now racing to expand their own reach. OpenAI has opened offices across Asia and Europe and warned in a recent statement that Chinese firms like Zhipu AI are working to entrench themselves in emerging markets before US alternatives arrive.

As OpenAI CEO Sam Altman put it in May: “We want to make sure democratic AI wins over authoritarian AI.”

But China’s focus on practical, low-cost AI applications—and its embrace of open-source models—may prove more attractive to smaller, budget-conscious countries looking to leap into the AI age.

The future of AI, it seems, will not be decided solely in Silicon Valley or Beijing—but in Riyadh, Johannesburg, Santiago, and beyond.