The United States and China are locked in a competition to lead the world in artificial intelligence, a technology that drives economic productivity, military strength, and global influence. For years, American companies like OpenAI and Anthropic set the pace, while Chinese firms were seen as followers. That dynamic began to shift in 2025, when Chinese developers released open-source models that matched US performance at a fraction of the cost. These models, such as Alibaba's Qwen and Moonshot's Kimi, quickly gained users worldwide, including in Silicon Valley, where startups adopted them to cut expenses.

Washington responded by restricting exports of advanced Nvidia chips to China, hoping to slow its AI progress. Instead, Chinese companies adapted by optimizing their models for domestic hardware and focusing on efficiency. By late 2025, Chinese open-source models had been downloaded more than 540 million times, and US firms like Airbnb were using them. This backdrop of export controls, cost competition, and rapid innovation set the stage for the recent narrowing of the performance gap between US and Chinese AI systems.