A recent study suggests China is swiftly narrowing the artificial intelligence (AI) technology gap with the U.S., as reported by the South China Morning Post (SCMP) on Tuesday.
Stanford University’s Human-Centered Artificial Intelligence (HAI) released its AI Index report, revealing that Chinese AI models have nearly caught up to their U.S. counterparts in key AI performance metrics. In particular, the gap in large-scale Massive Multi-task Language Understanding (MMLU) assessments between Chinese and U.S. models has dramatically shrunk from 17.5 points two years ago to a mere 0.3 points by the end of 2024.
Chinese tech giants, including Alibaba, ByteDance, Tencent, DeepSeek, and Zhipu AI, made significant strides in the global AI market last year. DeepSeek’s V3 model garnered attention for its high performance using minimal computational resources. Meanwhile, Meta faced controversy over allegations of manipulating benchmark scores for its Llama 4 model.
Despite China’s progress, the U.S. maintains its lead in AI model development. U.S. firms and institutions rolled out 40 major AI models last year, compared to China’s 15 and Europe’s 3. The U.S. also dominates in AI investment, with private sector funding reaching $109.1 billion, dwarfing China’s $9.3 billion. This disparity is partly due to U.S. government restrictions on domestic investments in Chinese AI startups.
However, China continues to make strides in AI research and patent applications. As of 2023, it leads globally in AI-related academic papers (23.2%) and patents (69.7%).