China Closes the AI Gap: How Long Can the US Keep Its Lead?

For years, many analysts have described the United States as having a clear lead over China in artificial intelligence. New assessments now suggest that gap is narrowing quickly. Recent reports, including work from Stanford’s Institute for Human-Centered AI, say China is catching up across multiple indicators: research output and citations, patents, benchmark performance, and the rollout of large-scale models.

Several forces sit behind this shift. Beijing continues to channel heavy public investment into chips, cloud infrastructure, and national AI platforms, while also rolling out detailed rules for generative AI services. By contrast, the US is bogged down in political fights over safety, regulation, and industrial policy, making it harder to sustain a coherent long‑term strategy. Some experts argue that AI is already a genuine “bipolar” system, where neither Washington nor Beijing can claim unambiguous dominance.

This changing balance matters far beyond the two countries. For mid‑sized economies like Japan, dependence on US and Chinese cloud and model providers is deepening. Policymakers in Tokyo will increasingly have to decide which ecosystems to plug into, how to diversify, and how to safeguard technological sovereignty and national security in a world where AI power is more evenly distributed.