American leadership in artificial intelligence may hinge not on research prowess alone but on adoption, according to a new memo from the Special Competitive Studies Project (SCSP), a nonprofit and nonpartisan initiative that recommends ways to strengthen America's long-term competitiveness in AI. The memo, titled The Open Frontier: How America Wins the AI Diffusion Race, contends that open models will play a decisive role in determining which technology ecosystem becomes the global default for governments, companies, and developers.
The stakes are illustrated by a historical parallel. Although the United States led the world in wireless research, Huawei won the diffusion race and connected much of the globe by out-competing on cost, financing, and speed of deployment. China is now implementing a similar strategy with AI, the memo notes. Its capable, low-cost, downloadable models are drawing developers, and developer adoption has shifted measurably toward Chinese open models. The advantage, according to the memo, comes not from openness itself but from the adoption it enables.
While there are legitimate trust and security concerns regarding Chinese open models, the American response should be to strengthen its own open model ecosystem rather than to restrict openness, the memo argues. Currently, leading U.S. frontier laboratories monetize access through subscriptions, enterprise products, and managed platforms. But a strong American strategy should also support open model developers and the researchers, startups, and businesses that build on their work.
The memo outlines three lines of effort. The first, Build the AI Ecosystem America Can Win, calls for expanding access to compute, capital, and shared infrastructure so that researchers, startups, and governments can inspect, adapt, and trust American models. The second, Turn Leadership into Adoption, recommends funding tools that demonstrate a model works for its purpose, aggregating and sharing working deployments, and measuring cost, time to complete, and mission outcomes to accelerate adoption of American AI across society. The third, Govern American AI Without Making It Inaccessible, urges policymakers to govern risk based on what models can do, how they are used, and their impacts and outcomes, not simply whether their weights are public.
The implications extend beyond Washington. For U.S. developers and startups, a thriving open ecosystem could lower barriers to building on advanced AI and reduce dependence on closed, subscription-based platforms. For governments and global buyers, the availability of trusted American open models could offer a credible alternative to Chinese systems, shaping which standards and norms take root worldwide. For the broader AI industry, the memo signals that the next phase of competition will be defined by diffusion—how quickly and widely a technology is adopted—rather than by research leadership alone.
SCSP's recommendations arrive as policymakers and industry leaders debate how to balance innovation, security, and openness in AI. The memo's central warning is that without a deliberate strategy to support open model development and adoption, the United States risks repeating the wireless experience, ceding the global default to a competitor that prioritizes cost and speed over the full stack of American strengths. Readers can learn more and read the full report at scsp.ai.
