The U.S. and China both recently unveiled their major Artificial Intelligence (AI) development plans recently. Experts view the U.S. as focusing on innovation and computing power, while China is focusing on applications.
Several studies have compared the AI capabilities between the U.S. and China:
- Stanford AI Index 2025: The U.S. developed 40 influential AI models in 2024, while China developed 15.
- HumanEval (code generation benchmark): The performance gap between the U.S. and China narrowed from 31.6 percent to just 3.7 percent.
- Insikt Group: Chinese generative models are only 3 to 6 months behind their U.S. counterparts, though structural disadvantages remain.
- In computing power, the U.S. controls 75 percent of the world’s AI supercomputing resources, while China holds only 15 percent. Even with plans to boost domestic computing capacity by 50 percent by 2025, China faces bottlenecks due to restrictions on high-end chips.
- Stanford AI Index 2025: The U.S. private-sector AI investment reached $109.1 billion in 2024 – 12 times that of China. Investment in generative AI alone exceeded $25.4 billion, more than the combined total of China, Europe, and the UK.
Looking ahead, experts say that while China benefits from massive user data and strong state support for smart cities and governance applications, it remains limited by U.S. export controls on chips and its reliance on overseas training for top AI talent. China’s core algorithms and foundation models still lack originality. As one expert noted, “China’s AI advances rapidly at the application layer, but basic research is still largely imitative.”
Source: Epoch Times, August 3, 2025
https://www.epochtimes.com/gb/25/8/2/n14565961.htm