U.S. vs. China: Who’s winning the AI race?

The U.S. and China are likely to develop complementary AI ecosystems rather than produce a single winner.

• U.S. companies such as Anthropic and OpenAI are spending heavily on frontier models, human-level intelligence and possible superintelligence, while China prioritizes industrial applications.
• China’s lower-cost strategy benefits from inexpensive labor and power, and its models use fewer tokens to address limited access to leading-edge computing capacity.
• Chinese companies are applying AI to autonomous driving, robotics, manufacturing and logistics, building on the country’s large industrial base and hardware capabilities.
• Open-weight Chinese models can be hosted locally by providers such as Amazon Web Services and Microsoft Azure, allowing U.S. customers to use them without sending data to China.
• Frontier U.S. models may retain premium work where a 10% to 15% performance advantage matters, while cheaper Chinese models could serve lower-stakes tasks; model-routing companies such as privately held OpenRouter already switch between systems based on cost and accuracy.

Takeaway: Investors should expect continued volatility and leadership changes rather than a definitive winner; U.S. AI infrastructure spending may eventually give way to growth tied to installed capacity, benefiting companies such as Amazon Web Services and potentially Meta if chip prices decline.