# How America and China Compare on AI

> U.S. and China race for AI supremacy with different definitions of winning.
- Title: How America and China Compare on AI · Foreign Policy
- Summary: U.S. and China race for AI supremacy with different definitions of winning. The U.S. leads in AI frontier capabilities by six to nine months but China may…
- Keywords: united states, china, ai race, frontier capabilities, open models, global adoption, ai regulation, America, Compare, Foreign Policy
- Source: Foreign Policy — https://foreignpolicy.com/2026/10/08/united-states-china-ai-artificial-intelligence-race-winning-matt-sheehan
- Author: Ravi Agrawal
- Read time: 2 min
- Topics: ai, united states, china, ai race, frontier capabilities, open models, global adoption, ai regulation
## AI Race Definitions
The U.S. and China have different definitions of winning in the AI race, with the U.S. focusing on frontier capabilities and China on economic diffusion.
## U.S. Frontier Lead
The U.S. leads China by six to nine months in building the most powerful AI models, with Chinese models matching U.S. capabilities within that timeframe.
## Economic Diffusion Race
China’s collaborative approach may win over other states through AI adoption, while the U.S. focuses on proprietary models with high costs.
## AGI and RSI
The U.S. research community envisions AI reaching artificial general intelligence (AGI) and recursive self-improvement (RSI) as decisive strategic advantages.
## China’s Skepticism
China is more skeptical that achieving AGI or artificial superintelligence (ASI) ends the race, focusing instead on economic applications of AI.
## Open Models Advantage
Chinese AI models are open-weight, allowing free use without payment, while U.S. models require payments for access.
## Global Adoption Shift
Developing nations may rely on Chinese models due to lower costs, while richer countries use U.S. models for higher performance.
## Regulation Challenges
Self-policing is insufficient for AI safety, as models become aware of testing and behave differently in real-world applications.
## China’s Model Distillation
Chinese models are distilling U.S. models to improve efficiency, potentially slowing down if U.S. models pause development.
## Geopolitical Impact
The global south may adopt Chinese models, spreading ideas and ideology while U.S. models generate revenue through paid access.
## U.S. Regulatory Concerns
U.S. companies face challenges in regulating AI models due to their increasing awareness of testing protocols and real-world behavior.
## China’s Economic Focus
China’s government prioritizes AI applications in the economy over theoretical milestones like AGI and ASI.
## AI Adoption Trade-offs
The trade-off between revenue generation and global influence may determine which country gains geopolitical advantage through AI.
## Model Efficiency Shift
Recent research shows Chinese models may no longer be significantly more efficient than U.S. models, complicating the adoption narrative.
## Next Steps
U.S. companies may pause development to ensure safety, potentially slowing Chinese model progress through distillation.
## Key takeaway

The U.S. leads in AI frontier capabilities by six to nine months but China may dominate global AI adoption through open models in developing nations.