Kazakhstan and Uzbekistan may be pressured to choose between the two powers they have spent decades balancing.

Since their independence 35 years ago, Kazakhstan and Uzbekistan have spent significant time practicing strategic balancing in their foreign policies. They have sought to counterbalance the interests and demands of their large neighbors, while courting would-be partners elsewhere. In doing so, they have cultivated constructive relationships with the United States, Europe, Turkey, the Gulf states, Japan, South Korea, Russia, and China. By alienating none and partnering with all to secure trade and connectivity, the strategy has served them well so far.

Now, however, advances in semiconductors, artificial intelligence, and digital technologies appear to be testing whether soft multi-vector balancing can survive. The United States and China are developing competing international frameworks that increasingly seem bent on partitioning access to technologies that promise to drive the next generation of economic and geopolitical power.

Washington’s Pax Silica initiative features access to trusted supply chains of semiconductors, artificial intelligence, critical minerals, advanced manufacturing, logistics, energy, and digital infrastructure. To secure exclusive hardware access and technology sharing, membership in this coalition of like-minded countries may soon come with strict requirements on principles such as data privacy and ethical AI development.

China’s World Artificial Intelligence Cooperation Organization (WAICO) may sound like a counterproposal on the surface, but it is instead an open-access, inclusive grouping of countries seeking broad international membership. Joining WAICO includes technology sharing, as well as open-source development and assistance for building AI capabilities. For countries seeking to develop and advance their own technological capabilities, both propositions seem attractive. One, however, is increasingly being represented as incompatible with the other.

Central Asia’s Choice Between China and the United States

Reuters reported in August that the United States was preparing to tell countries that seek to participate in both Pax Silica and WAICO will have to choose between the two. Though not fully articulated, the US policy direction views participation in China’s technology ecosystem as a security vulnerability for high-end US technological components and as an incompatible, competing vision for artificial intelligence. This discrepancy creates a particularly difficult challenge for Kazakhstan and Uzbekistan, as both seek to participate in both initiatives.

Kazakhstan joined Pax Silica in June 2026, then joined the China-led WAICO framework several weeks later, in July. Uzbekistan seeks similarly to belong to both organizations. Consistent with how they traditionally manage international relationships, such instincts are understandable. But this may be a case in which multi-vector diplomacy runs into a hard technological wall.

Pax Silica and WAICO are not hi-tech embodiments of West and East that skilled actors can bridge. The two represent very different approaches to an emerging technological order. Pax Silica extends access to trusted supply chains and reduced dependence on potentially coercive suppliers. The US framework includes all-encompassing access to semiconductors, connectivity, advanced manufacturing, logistics, and energy. Beyond AI access, Astana and Tashkent are drawn to American and allied technology, investment, supply chains, and markets.

The WAICO model is different and oriented towards design and development. Its approach emphasizes access to AI, open-source technologies, capacity building, and reducing the technological divide between advanced and developing countries through the roll-out of shared platforms. China’s proposal is attractive to developing countries, as it promises access to technology without the political and security conditions requisite to Western techno-systems.

Central Asia Can’t Have Its Tech Stack Both Ways

For Kazakhstan and Uzbekistan, both notions are attractive. Pax Silica can provide access to trusted technology and Western capital while reducing vulnerabilities associated with Chinese-controlled infrastructure. WAICO can provide access to comparatively inexpensive Chinese technology, open-source AI, technical assistance, and a network of developing countries.

The liabilities and overlaps are equally real. A close technological relationship with China will create long-term dependencies tied to equipment and operating systems. A similarly close relationship with the United States, in exchange for high-performance computing equipment, will limit the inclusion of Chinese equipment and technology that Central Asian governments have already purchased and mainstreamed into their economies. The issue is not just infrastructure, but also political economy.

The development of the Middle Corridor is a prime example. More than an exercise in improving connectivity across ports, rails, and highways, the route also depends on adopting data, software, and digital systems to move freight competitively. Relying on and integrating these disparate components changes the meaning of each choice, especially as they relate to one another.

When a country chooses a cloud platform, telecommunications provider, data system model, AI structure, or digital freight handler, it is not necessarily making a geopolitical decision in that moment. But collectively, those decisions produce an integrated network with embedded, standardized operating protocols. Systems become interoperable with personnel trained to operate specific technologies. Data aggregates and becomes siloed within these systems. The cost of making changes or switching components becomes enormous.

In effect, joining a technology framework and adopting a specific supply chain can drive long-term decisions about economic alignment and digital sovereignty. What might initially have seemed as a technical or commercial stand-alone choice can become a Trojan horse for future procurement, interoperability, or any notion of strategic digital autonomy. Kazakhstan in particular has already navigated a number of such decisions.

One past example of how dependencies can develop is what Astana faced at the telecom level: adopting 5G systems and deciding whether to rely on Huawei or European suppliers such as Ericsson and Nokia for critical telecommunications infrastructure. The decision initially may have seemed about which equipment was cheaper versus technically superior. However, once a specific telecommunications vendor’s equipment is adopted and rolled out, replacing it can be prohibitively expensive and disruptive. Not inherently a geopolitical choice when purchased, after thousands of such decisions, the resulting digital ecosystem evolves and becomes difficult to replace.

Kazakhstan’s experience with the US-led Blue Dot Network illustrates the challenge. Blue Dot is generally a quality-assurance framework for project financing with high infrastructure standards and project certification. Not an exclusive technology alliance, Blue Dot can coexist with other systems and technologies. Astana expressed interest in joining Blue Dot but never did, while continuing to accept Chinese infrastructure and technology as part of Beijing’s Digital Silk Road (DSR) initiative, which offered telecom systems, AI capabilities, cloud computing, e-commerce, and other high-tech capabilities. Under Blue Dot, the old multi-vector model was possible; Pax Silica illustrates why it may now be breaking down and presents a new challenge.

Pax Silica’s Central Asian Sand Trap

Pax Silica, however, presents a potentially different challenge. If Washington limits access to advanced computing and technology to participation only behind a firewalled technology system, Astana’s existing Chinese equipment and infrastructure could limit the integration of new US technology into existing systems, depending on Pax Silica’s security, technology, and financing restrictions.

This is a serious, cross-cutting issue that impacts all sectors of a country’s economy. Digital infrastructure increasingly connects and becomes embedded in physical systems that form a common network, with information about cargo, shipping, receivers, and transactions part of the same network. Thus, as a technology provider becomes rooted in Kazakhstan or Uzbekistan’s digital infrastructure, it becomes difficult, if not impossible, to separate commercial technology from questions of data security and strategic dependence.

For Central Asia, this problem is especially acute because the region has historically depended on terrestrial digital connections running through Russia and China. That is beginning to change as new subsea connectivity is being developed across the Caspian Sea and as satellite-based systems such as Starlink arrive. Both offer Central Asian governments alternatives to traditional geographic chokepoints and greater control over how their data moves beyond their borders. These new openings make the choice between technological ecosystems more critical.

Astana and Tashkent continue to wade into difficult waters, signing several significant digital development agreements with China in addition to joining WAICO. One such project is Kazakhstan’s massive Data Center Valley, a one-gigawatt data center complex that envisages creating hyper-scale data centers with Chinese-supplied hardware to develop supercomputing capacity, AI laboratories, and next-generation research institutions. 

Alongside Chinese supplier Huawei, US companies Nvidia and Firebird are also $10 billion core participants in the same massive AI complex. Today, Kazakhstan can place Chinese and US technology into the same development, but if Pax Silica imposed hardened ecosystem boundaries, that flexibility would likely disappear.

This juxtaposition of clashing techno-systems is evident across the region’s evolving Middle Corridor. What began as decisions about railways, ports, and trade routes linking Central Asia to Azerbaijan, Georgia, Turkey, and Europe is increasingly about more than that. The corridor’s competitiveness will depend on digital systems, including customs, cargo tracking, payments, logistics, and potentially AI management of the entire system.

The physical corridor and the digital corridor are merging into one system. Whoever supplies the digital architecture will therefore influence more than technology. It will have a role in determining how the region’s trade operates. But Washington and Brussels should be careful, as Central Asian governments understand the risks of dependence precisely because they live between Russia and China. Their instinct is to diversify. They want US technology, Chinese investment, European markets, Russian security ties, and connections to the Gulf and Asia. 

The problem is that today’s emerging technology systems are cumulative. Membership leads to standards. Standards influence procurement. Procurement guides infrastructure. Infrastructure establishes dependence. This is the root of Washington’s strategic concern over Beijing’s technological penetration. Central Asian governments find themselves in a Pax Silica sand trap, seeking access to next-generation AI microchips while already committed to Chinese infrastructure. To abandon these latter investments requires Washington to make a compelling case.

If Pax Silica becomes a genuine alternative technological ecosystem, offering competitive financing, reliable technology, investment, and long-term commercial partnerships, then Kazakhstan and Uzbekistan will have strong reasons to pivot. If Washington offers security arguments without comparable commercial alternatives, the choice becomes harder. This is ultimately the test for Pax Silica—and for the Central Asians who have tended to avoid such either-or decisions in the past.

The United States has to offer something better if it expects the Central Asians to reject Chinese technology simply because Washington regards it as strategically risky. Building a wall to defend against digital China risk will require major choices for Central Asians and incentives from the United States. Competitive access to American and allied technology, financing, investments, and partnerships in data centers, telecommunications, and cloud infrastructure. It means helping Central Asian countries develop their own AI expertise rather than simply asking them to buy American, and demonstrating that US engagement will continue beyond the next election or administration.

The First Test for Central Asian Multi-Vector Diplomacy

Events are moving rapidly, however, and the technology dilemma no longer appears hypothetical. Kazakhstan became the first country to join both Pax Silica and WAICO, signing the US-led declaration in June and the Chinese-led WAICO agreement several weeks later. Uzbekistan subsequently sought to follow the same path, formally asking Washington to join Pax Silica after becoming a WAICO member. 

These dual enrollments demonstrate both sides of the emerging problem. Central Asian governments are again attempting to avoid a choice between Washington and Beijing. But the Trump administration is increasingly signaling, including at the UN General Assembly, that participation in China’s system could become incompatible with, or disqualify one from, access to the most sensitive elements of the American system. The result is a test of whether multi-vector diplomacy can survive when technological ecosystems become mutually dependent, and security boundaries grow harder to cross. 

Can the nations remain digitally nonaligned as advanced computing and its associated supply chains become increasingly integrated? Central Asia may provide the first important test case for this geoeconomic challenge.

The United States needs to understand that Central Asian governments are not looking for a new hegemon but for options. Washington must not interpret the Pax Silica-WAICO confrontation as just another US-China skirmish. So far, Central Asians are seeking to preserve multi-vectoral flexibility amid competing technological systems. But these systems are becoming less interchangeable as they are deployed.

Central Asia’s multi-vectoral strategy was developed in a world in which countries could counter-balance competing partners. If Washington’s exclusivity policy becomes formal, Kazakhstan and Uzbekistan may have little choice. But Washington also needs to appreciate what it is asking them to give up, particularly as China is a major and critical trading partner. For Kazakhstan and Uzbekistan, this is not a simple choice between two competing offers, but a decision between technological ecosystems with consequences that will last for decades. In the end, this process is testing whether technological nonalignment will still be possible in the future.

The competition is not simply over who supplies the chips, but over who creates the entire digital ecosystem, including its long train of standards, practices, training, and structures that accompany those technologies. The Huawei-versus-Ericsson/Nokia debate was a warning. Technology choices that appear manageable today can easily become tomorrow’s critical dependencies. Pax Silica must therefore be more than a political declaration. To attract countries that have already adopted Chinese technology, the United States will need to offer a credible technological and commercial alternative.

The objective should not be keeping Kazakhstan and Uzbekistan out of WAICO. Instead, it should ensure they can enter the AI age with digital sovereignty, security, and a genuine choice of partners. For countries whose overriding strategy has been to preserve their freedom to maneuver, keeping that choice may be the most important outcome of all.

About the Author: Eric Rudenshiold 

Dr. Eric Rudenshiold served for four years as the US National Security Council’s Central Asia Director under presidents Donald Trump and Joe Biden. He is currently a senior fellow at the Washington office of the Caspian Policy Center. Prior to serving at the NSC, Dr. Rudenshiold served in a variety of capacities for the US Agency for International Development from 2007 to 2018, including acting roles as deputy assistant administrator, country office director, director for South and Central Asia, senior officer in charge of Central Asia, and senior advisor.

The post Central Asia’s Technology Trap appeared first on The National Interest.