What Happens If the World Splits Into Two Technology Systems?

A phone, car or AI service can look like one product. Underneath it may depend on chip designs from one country, manufacturing tools from another, fabrication concentrated in Taiwan, minerals processed elsewhere, and software distributed around the world. The politics of technology now reach into all those links.

The phrase “two technology systems” suggests a clean division: one led by the United States and its partners, another by China. That is a useful shorthand for strategic competition, not a literal description of how technology works today. The evidence shows selective controls, subsidies, and efforts to build alternatives—not a completed global divorce. A split, if it deepens, would be uneven: chips and AI infrastructure may separate faster than ordinary consumer goods or the basic protocols that let networks communicate.

The question is not simply who would win. It is what changes when access to technology depends increasingly on where it was designed, made, hosted, or governed.

The split is already visible—but not complete

The clearest dividing line runs through advanced semiconductors. In December 2024, the U.S. Commerce Department announced controls covering 24 types of chipmaking equipment, three categories of software tools, high-bandwidth memory, and 140 additions to its Entity List. Washington said the measures were intended to restrict China’s ability to produce advanced chips for military applications. That is the government’s stated rationale; the announcement itself does not establish how effective the controls will be over time. [1]

China has also used export controls. A Ministry of Commerce notice issued the next day prohibited exports of covered dual-use items to U.S. military users or for military uses, and said exports of gallium, germanium, antimony and superhard materials to the United States would, in principle, not be permitted. It also tightened checks on graphite exports. This is a specific measure, not evidence that China stopped exporting all these materials to every destination. [2]

Each move illustrates a wider shift: governments are treating some commercial technologies and inputs as matters of national security. That does not mean all trade is being cut off. A Congressional Research Service review in 2025 noted that U.S. controls had restricted some advanced technologies while other parts of the semiconductor supply chain remained open to China. [3]

The result is a patchwork of restrictions, licences, subsidies and carve-outs. A company may still sell in a market, but face new conditions on which chips it can supply, what equipment it can maintain, or where it can expand a factory. Legal boundaries can change faster than physical supply chains can be rebuilt.

Why chips are the pressure point

A chip is not a single-country product. Design software, intellectual property, manufacturing equipment, specialty materials, fabrication, packaging and testing are distinct stages. The OECD’s 2025 mapping of the semiconductor value chain describes it as globally distributed and tightly interconnected, with critical inputs concentrated in particular regions and economies specialized in different segments. It warns that disruption in a leading economy can cascade into shortages for downstream industries. [4]

That concentration is especially consequential for Taiwan. TSMC’s 2024 annual report lists its largest set of fabrication facilities in Taiwan, alongside operations in China, Japan and the United States. Its report also describes surging AI demand and record revenues in 2024. These are company-reported facts; they show both how central one manufacturer is to global production and how much activity is already distributed across borders. They do not mean overseas sites can immediately duplicate every capability at home. [5]

Governments are trying to reduce risks by encouraging more local production. The United States’ CHIPS Act, for example, provides incentives for domestic semiconductor investment. Its final security rules also restrict recipients from materially expanding advanced-chip manufacturing in countries of concern for ten years, subject to the rule’s scope and exceptions. Such policies can add capacity and provide insurance against disruption. They also tie investment decisions more closely to political blocs. [6]

But building a fab is not the same as replacing a whole ecosystem. It takes capital, specialized workers, reliable electricity and water, suppliers, customers, and years of process learning. OECD analysis says diversification requires substantial investment, skilled labor and infrastructure. **Analysis:** if governments pursue resilience by duplicating every stage inside friendly borders, consumers and firms may pay for redundant capacity while still relying on external partners for some inputs. [4]

What two systems would mean for business and people

For companies, the immediate cost is complexity. A product sold worldwide might need separate chip configurations, cloud hosting, software updates, compliance reviews, or supplier networks depending on destination. If rules diverge, businesses could need separate certifications and technical teams. Smaller companies, which cannot afford parallel operations as easily as global giants, may find some markets harder to enter. These are plausible effects of regulatory and technical divergence, not guaranteed outcomes for every firm.

For consumers, the consequences may show up as higher prices, fewer choices, delayed features or inconsistent services across borders. A firm that cannot legally provide the same software, AI model or cloud service everywhere may offer a reduced version—or withdraw. Restrictions can also limit access to useful technology and research. At the same time, public support for domestic capacity could create jobs and make some supplies less vulnerable to a crisis. Security and efficiency can pull in opposite directions.

The costs are difficult to forecast. The IMF has summarized estimates that greater international trade restrictions could reduce long-run global output by as much as 7 percent, or about $7.4 trillion in the dollars used for its calculation. That is a scenario-dependent estimate of broad fragmentation, not a prediction that a technology split alone will cause that loss. IMF researchers identify trade, capital flows, technology diffusion and public goods as channels through which fragmentation could impose costs. [7]

Trade data also counsel against treating “two blocs” as a settled fact. A WTO working paper analyzing goods trade through February 2024 found early signs of friend-shoring after Russia’s invasion of Ukraine, but no increased regionalization after the pandemic or the invasion. It reported that trade between hypothetical East and West blocs grew more slowly than trade within them, while also distinguishing U.S.–China decoupling from wider geopolitical alignment. The authors call this work in progress, not an official WTO position. [8]

Not just chips: rules, data and materials

Technology systems are made of rules as well as machines. The European Union’s AI Act sets a risk-based legal framework for AI in the EU, with obligations taking effect in stages. The OECD’s AI Recommendation, amended in 2024, offers a common intergovernmental set of principles, including human rights, transparency, safety and accountability. These examples show that different jurisdictions can regulate differently while still drawing on shared vocabulary. A different rulebook does not automatically mean incompatible software or a separate internet. [9] [10]

The United Nations’ 2024 Global Digital Compact calls for an open, global, interoperable and reliable internet, while supporting international cooperation on data governance and AI. Its existence is evidence of an effort to preserve common ground, not proof that governments will resolve disagreements over surveillance, privacy, censorship, security or data access. [11]

Critical materials add another point of leverage. OECD analysis of raw-material supply chains says the three largest producing countries accounted for 90 percent of rare-earth-element production in 2023. Extraction and processing are concentrated for geological and industrial reasons; restrictions can ripple through international markets. That concentration creates exposure in both directions: countries that rely on foreign processing face supply risks, while producers can face lost revenue and customers may seek alternatives. [12]