China is fast catching up to the US in AI, and its progress is seen in the accepted papers at NeurIPS, a top AI conference.
The latest data from the Neural Information Processing Systems (NeurIPS) 2025 conference reveals a remarkably tight race between the world’s two AI superpowers. According to statistics compiled by papercopilot.com, the United States leads with approximately 2,450 accepted papers, while China follows closely with 2,370 papers—a gap of less than 5% that underscores China’s rapid advancement in artificial intelligence research.
This near-parity represents a significant shift in the global AI research landscape. NeurIPS, one of the most prestigious venues for machine learning and computational neuroscience research, has long been considered a bellwether for innovation in the field. The conference’s acceptance statistics provide a snapshot of where cutting-edge AI research is being conducted and which countries are leading the charge.
The data shows a steep drop-off after the top two positions, with the United Kingdom placing third at approximately 500 papers—less than a quarter of either the US or Chinese totals. Singapore, Germany, and South Korea round out the next tier, each contributing between 200-400 papers, while other nations trail significantly behind.
There have been other indications that China is fast catching up to the US in AI. Chinese open-source models have passed US-open source models in popularity. NVIDIA CEO Jensen Huang had earlier said that 50% of AI researchers were Chinese, and 70% of AI patents were being granted to China. China is also adding electricity capacity a lot faster than the US, which could help its AI efforts in the coming years.
The implications for the global AI industry are substantial. Research output at conferences like NeurIPS often foreshadows commercial applications and technological breakthroughs. China’s strong showing reflects its sustained investment in AI research infrastructure, talent development, and academic institutions over the past decade. The country has made AI development a national priority, with significant government funding and initiatives aimed at becoming a world leader in the technology by 2030.
For businesses and policymakers, these numbers serve as a reminder that the AI race is truly global and increasingly competitive. The near-equal footing of US and Chinese researchers at NeurIPS 2025 suggests that innovation in artificial intelligence will likely continue to emerge from both nations in roughly equal measure, with profound implications for everything from semiconductor demand to regulatory frameworks and international collaboration in the field.