Measuring the US-China AI Gap

This report assesses China’s ambition to lead AI by 2030 across pillars such as funding, regulation, talent, diffusion, model performance, and compute, concluding that while China’s capabilities are rapidly improving—highlighted by milestones like DeepSeek’s R1, strong academia–industry collaboration, and growing open-source adoption—it is unlikely to sustainably surpass the US in the near term. The analysis finds US private investment and talent remain advantages, China’s regulations slow public-facing innovation, and semiconductor constraints persist despite notable domestic progress and state-backed funding. The report emphasizes that diffusion, not just innovation, will determine economic and geopolitical gains, notes increasing reliance on UGC data and risks from espionage and model distillation, and recommends Western stakeholders strengthen IP protections, export-control due diligence, and global talent recruitment while closely tracking Chinese AI developments.