A policy analysis outlines export-control and 'Tech West' strategy for rebuilding AI infrastructure resilience outside China.
Two facets of American AI policy are often treated as contradictory, when in reality they are two sides of the same coin: Washington wants to aggressively export US-designed chips and cloud services to make American hardware the backbone of the global intelligence economy, while simultaneously keeping Chinese AI labs compute-constrained by refusing to sell them the high-end chips needed to reliably offer AI services to global publics. A recent AEI report argues that there is no contradiction here. American compute should be the easiest in the world to buy and the hardest to steal. Washington can—and should—pursue a policy of deliberately exporting high-end compute to most countries capable of installing it, with one major exception.
Some China hawks worry that shipping advanced chips abroad creates obvious enforcement loopholes. Chips installed in data centers in Malaysia or Thailand are often rented remotely by customers in Beijing or Hangzhou. Racks of servers sitting in Southeast Asia may effectively amount to Chinese compute by another name. This is a real problem that Congress is working to address, but it is not an argument for keeping every GPU at home, as some have begun to advocate.
Computing capacity is an important element of national power, and the United States should build as much AI infrastructure domestically as it practicably can. But there are limits to how quickly American communities will accept new data centers and their requisite sources of power generation, transmission, and storage. Political opposition to large data centers is growing, environmental permitting remains difficult, and grid interconnection can stretch on for years.
Other countries have the land, energy, and political appetite needed to build data centers—resources the United States now struggles to muster in abundance. As frontier AI systems grow larger, it is becoming clear that no single country can host sufficient compute to perform cognition at planetary scale. For this reason, Washington should welcome the division of labor among a wider "compute coalition."
If the United States refuses to help trusted countries build AI infrastructure, those countries will necessarily look elsewhere. China is trying to build its own globally competitive stack of chips, chipmaking equipment, networking gear, and cloud services. While the US-led chipmaking ecosystem still enjoys an enormous qualitative and quantitative lead over China's limited production, this American advantage will not last forever. The goal should not be maximum concentration of compute in the United States or maximum diffusion everywhere else. Rather, Washington should seek to maximize the deployment of compute inside a US-aligned global technology ecosystem—a "Tech West" spanning the United States and trusted partners that build, host, and govern the infrastructure for advanced AI.
The time to build this system is now, while the United States remains the world's sole provider of high-end compute. By attaching specific requests to large-volume exports of AI chips, Washington can spend its waning influence to design a global computing ecosystem that advantages the United States and disadvantages China.
Building the Tech West requires deft diplomacy. Asking for too much risks alienating partners or pushing them to explore Chinese alternatives, while asking for too little risks defeating the purpose of a coalition.
A middle-ground policy is possible. Large deployments of advanced American chips abroad should come with meaningful requirements governing remote access and customer verification for computing services. Those rules should be designed to make legitimate overseas deployment easy. Trusted governments, American cloud providers, and vetted foreign firms should have a clear path to buying enormous quantities of US hardware and quickly standing up computing capacity for AI inference and training. The point of export controls is to deny strategic technology to adversaries, not to disadvantage American companies in global markets.
Membership in this ecosystem need not depend on participation in a formal alliance. Countries willing to enforce common rules against chip diversion or unauthorized cloud access should be able to participate.
America occupies an extraordinary position in the global AI supply chain. Most countries want to buy US chips and cloud platforms. This supply-demand crunch has handed Washington a generational opportunity to shape where AI infrastructure is built and who can access it—but the window to build a Tech West will narrow as Chinese alternatives improve. The United States should therefore spend its extraordinary leverage on building a globe-spanning network of American AI infrastructure while ensuring Beijing cannot covertly buy, borrow, rent, or steal access to the same.