US-China AI Competition Shifts to Ecosystem Statecraft

For much of the past several years, the debate over artificial intelligence has revolved around two foundational questions: Can American technology companies continue to innovate at the technological frontier? And can the United States government develop a strategy capable of preserving America's technological advantage over China? Those questions have now been overtaken by events.

The defining question is no longer whether China can compete at the frontier. It is whether the U.S. can adapt quickly enough to compete against an increasingly sophisticated Chinese innovation ecosystem that is advancing not only on model performance but also on cost, deployment, customization, financing, standards, developer adoption, and global reach. Washington increasingly finds itself responding to successive Chinese breakthroughs rather than shaping the competitive environment in which artificial intelligence develops. That should concern policymakers, technology executives, investors, and America's allies far more than the latest benchmark score or model release because the competition is rapidly evolving beyond individual companies and becoming, instead, a contest between competing innovation ecosystems.

The headlines surrounding DeepSeek, Moonshot AI's Kimi K3, Alibaba's Qwen family of models, Tencent's Hunyuan, Zhipu AI and MiniMax are often treated as separate stories. They are anything but. Viewed collectively, they reveal something far more consequential than the emergence of several successful Chinese AI companies. They demonstrate that China has cultivated a frontier AI ecosystem capable of repeatedly producing world-class capabilities across multiple firms. Whether those advances emerge through original innovation, engineering optimization, open-weight collaboration, or from distillation of U.S. models is increasingly beside the point. The larger strategic reality is that they are occurring across an ecosystem, while the U.S. continues to evaluate them one company at a time and often responds as though each breakthrough were an isolated event rather than evidence of a broader structural transformation.

Over the past several years, the U.S. has consistently underestimated China's commitment to long-term technological advancement and its ability to translate domestic industrial strategy into global competitive advantage. Whether the issue was rare earths, electric vehicles, robotics, semiconductors, or artificial intelligence, the analytical mistake has remained remarkably consistent. Washington has tended to evaluate China's progress company by company and product by product, often dismissing each advance as exceptional or unsustainable, while Beijing has pursued a patient strategy designed to cultivate the conditions under which an entire ecosystem could innovate and deploy simultaneously.

It is equally important to recognize that China's plans and long-term trajectory toward becoming a technology superpower were established years before the Biden administration's technology restrictions. Those measures may have influenced the direction and pace of Chinese innovation, but they did not create the underlying strategic trajectory. DeepSeek's January 2025 announcement compelled many observers to acknowledge a trajectory that Beijing had been articulating through industrial policies, successive Five-Year Plans, and national technology strategies for more than a decade. The breakthrough was not the strategy. It was evidence that the strategy was beginning to produce measurable impressive results.

We are entering an era of ecosystem statecraft

That broader approach is what I would describe as ecosystem statecraft: a form of strategic competition that seeks to shape the competitive environment within which technologies are developed, financed, standardized, deployed, and ultimately promoted and adopted. It integrates industrial policy, finance, innovation, global standards, university curriculum direction, state-supported developer ecosystems, diplomacy, and commercial expansion into a coherent national strategy designed to reinforce long-term technological leadership. Rather than competing company by company or technology by technology, ecosystem statecraft seeks to shape not just the technologies themselves, but the conditions under which they succeed.

Artificial intelligence simply happens to be the clearest manifestation of this broader strategy today. The same logic existed across China's approach to semiconductors, electric vehicles, batteries, robotics, telecommunications, renewable energy, critical minerals, digital infrastructure, and advanced manufacturing. AI is therefore not an exception to China's industrial strategy. It is its most sophisticated expression.

Viewed through that lens, the United States and China increasingly appear to be pursuing fundamentally different theories of victory. American policy has understandably emphasized preserving technological leadership through frontier innovation while slowing China's progress through export controls, investment screening, and restrictions on access to advanced computing. Those remain important tools and should continue to play a central role in America's competitive strategy. Beijing increasingly appears focused on shaping the ecosystem within which global technology competition occurs. Chinese AI firms are making their technologies easier to deploy, easier to customize, easier to integrate across multiple computing environments, and easier for developers, businesses, and governments around the world to build upon. In the long run, reducing friction throughout the technology stack may prove just as important as improving benchmark performance.

A race to 'addict' the rest of world to tech stack

President Xi Jinping's recent address to the World Artificial Intelligence Conference reflected this broader vision. By emphasizing international AI cooperation, governance, open-source development, and greater participation by developing countries, Beijing continued to position itself not simply as a producer of advanced AI, but as the architect of an alternative global technology ecosystem. Viewed together with China's efforts to strengthen domestic control over strategically important technologies while encouraging international adoption of its AI platforms, the strategy becomes increasingly clear: protect critical capabilities at home while expanding technological influence abroad. Or, as Commerce Secretary Howard Lutnick said during public debate over an Nvidia chip export ban -- echoing a talking point from Nvidia CEO Jensen Huang -- the goal is "addicting" the rest of world to a tech stack. But increasingly, it is not clear that the American stack is the one.

This broader strategy also helps explain why persuading countries to avoid Chinese AI will likely prove considerably more difficult than Washington's earlier campaign against Huawei and ZTE. Governments can regulate telecommunications infrastructure, but they have far less ability to determine which AI models, software libraries, and developer tools are ultimately adopted by millions of developers and integrated into commercial applications around the world. Increasingly, technology adoption is occurring from the bottom up as much as from the top down.

At the same time, many governments now evaluate both Washington and Beijing through a more pragmatic lens, balancing and hedging both countries, and making decisions based on security concerns as well as comparative affordability, financing, technological capability, local capacity building, and long-term economic opportunity. Countries ultimately adopt technology ecosystems from partners they believe will remain reliable, affordable, accessible, and committed to long-term engagement. Trust, financing, developer communities, standards, commercial partnerships, and diplomatic credibility have become competitive advantages in their own right.

What concerns me most, however, is not China's progress. It is the framework through which the U.S. increasingly debates its response. Too often, America's AI conversation is framed around the competitive interests of individual companies rather than the nation's long-term strategic interests. That is not a criticism of OpenAI, Anthropic, Nvidia, Microsoft, Amazon, Google, or any other company. Their responsibility is to maximize shareholder value and strengthen their competitive position. Governments have a different responsibility. Markets optimize for competitive advantage. Governments must optimize for national advantage. Those objectives frequently overlap, but they are not always the same.

America retains extraordinary advantages. Its universities remain unparalleled, its venture capital ecosystem remains unmatched, its semiconductor industry continues to underpin the global AI economy, and its frontier laboratories continue producing extraordinary breakthroughs. Yet history reminds us that technological leadership is rarely determined solely by who invents it first. More often, it is determined by who builds the ecosystem that everyone else ultimately chooses to join.

The question for the U.S. is therefore not simply whether American companies can continue building the world's most capable AI models. It is whether the United States can develop an equally coherent national strategy, one capable of surviving changes in administrations while combining technological innovation with trusted alliances, standards-setting, research, talent development, commercial partnerships, financing, and renewed international credibility. Once the competition becomes ecosystem versus ecosystem, success will depend on far more than whose models perform best. It will depend on whose ecosystem the world's developers, researchers, entrepreneurs, universities, businesses, governments, and investors choose to trust, adopt, and build upon. Given the totality of those metrics, the United States still has work to do.