The United States is ahead in the AI race against China overall, driven by advantages in chips, AI models, economic systems, and talent; however, victory should be judged by global adoption of AI technology, not merely technological supremacy.

Policy recommendations urge treating AI as critical infrastructure, supporting both frontier and open-weight models, increasing transparency about progress and setbacks, and moving toward a rapid, coordinated infrastructure designation and governance framework by the end of 2026.

Chinese models offer cost advantages, being 5-10 times cheaper per token, and are widely used by cost-sensitive buyers, which could pose security risks to the US software supply chain.

The AI adoption race is framed as a triathlon with three synchronized legs—innovation, cost, and trust—each essential for global standardization and broad adoption.

In trust, the US benefits from its values and open system, while China’s Great Firewall and CCP influence raise concerns about transparency and governance; public trust in AI has been wavering in the US amid data-center and disinformation issues.

Horacio Rozanski, CEO of Booz Allen, stresses urgency to accelerate and safeguard US leadership before China achieves broader global adoption of its AI ecosystem.

To win the adoption triathlon, the US should push forward technology, invest in lower-cost alternatives, and rebuild trust, using the Biden administration’s America’s AI Action Plan as a roadmap.

Innovation leadership remains strongest in the US through lithography and advances by Nvidia and TSMC, with frontier lab progress potentially 8-12 months ahead in models, though the gap is narrowing.