Artificial intelligence (AI) has already started reshaping the global trade landscape. It has the potential to become a powerful driver of inclusive, trade-led growth if economies invest in the right, enabling policies and cooperate to prevent fragmentation of the regulations governing the digital economy. But given AI’s broad application and its close ties to future national capabilities, including military ones, the competition to lead in the field has intensified to the extent that it now makes multilateral cooperation to craft global AI rules nearly impossible.
AI’s impact on trade
AI’s impact on trade can be grouped into three broad categories: Growth in AI-related supply chains; the commodification of trade-related processes; and rising international trade in AI-driven or AI-assisted products.
When it comes to supply chains, some countries are already experiencing a macroeconomic impact from their role in building AI infrastructure, most notably Taiwan. In September 2025, Taiwan’s exports reached US$70.2 billion, a 30.5% year‑on‑year increase and the highest monthly figure ever recorded — driven largely by AI‑related products. Taiwan’s economy is projected to grow close to 6% in 2025, far exceeding earlier forecasts, with officials explicitly citing the AI semiconductor boom as the main driver. South Korea is experiencing similar gains, positioning itself to dominate the AI memory chip market.
There is mounting evidence that AI is helping to reduce international trade frictions arising from customs clearance, cultural barriers, compliance with regulations, and, increasingly, changing tariff regimes. In the case of DHL, predictive analytics, powered by AI, reduced delays and cut costs by managing inventory levels, forecasting demand, and optimizing delivery routes. Small and medium-sized enterprises, which often underutilize trade agreements due to their lack of capacity to cope with complex trade policy documentation, could benefit disproportionately as AI lowers some barriers to entry by automating compliance tasks.
AI is also driving an international exchange of new products. Consider China’s AI toy market, which reached RMB24.6 billion in 2024 and is projected to grow to RMB29 billion in 2025. The impact of AI could be even more pronounced in the services trade. By lowering the cost of services, such as legal advice, AI could drive a surge in cross-border sales, potentially rivaling the effect on trade volumes in goods that resulted from China’s integration into the global trading system in the 1990s.
As with previous technological breakthroughs, the power of AI to transform economic capabilities makes the race to lead the industry a matter of significant geopolitical importance. Vastly different approaches to AI regulation have emerged. In its early evolution phase, the basic requirement of AI is already producing a bifurcating effect on the global economy, and in some cases, intensifying existing competition for resources.
Competition for the components of success
The outcome of the race to lead in AI will largely be determined by access to five key elements of the AI ecosystem: semiconductors, data, electricity, water, and critical materials.
Semiconductors
The semiconductor industry has become the most politicized sector in the global economy. From China’s US$100 billion superfund aimed at stimulating domestic design and fabrication, to the US CHIPS and Science Act aimed at reshoring semiconductor fabrication, the list of measures that impact the industry is long. In 2024, integrated circuit (IC) exports exceeded US$1 trillion, accounting for an estimated 4% of world trade in goods. The stated desire of both China and the US to achieve domestic self-sufficiency in semiconductors could well result in a stagnation of cross-border trade in semiconductors going forward. We could see production systems starting to bifurcate.
Data
AI models rely on massive, diverse, high-quality datasets to learn patterns. Those controlling the richest datasets can train the most capable models, provided they also have sufficient computing power. Data could therefore become a highly tradable commodity in the international marketplace. However, there are several issues pertaining to the trade of data. Most fundamentally, in property-owning societies, is establishing ownership: To whom does it belong? Does collecting data convey ownership? As the value of data is being realized, barriers have started to be erected to prevent cross-border flow of data. The European Union’s General Data Protection Regulation (GDPR) and China’s National Security Law are examples.
Electricity
AI already consumes as much electricity as a country the size of the UK. By 2030, its usage may be of an order of magnitude similar to Japan’s entire electricity consumption. Being able to accommodate such an increase in demand will require considerable infrastructure built. The ample availability of cheap energy is a key driver of data center locations, and the Middle East is carving itself out a competitive position in AI infrastructure as a result. Countries are scaling back environmental, social, and governance (ESG) regulations in response to the pressing urgency of not falling behind in the AI race.
Water
Water availability is becoming a key factor in deciding the locations of data centers. Protests are becoming more frequent in drought-affected areas where companies are seeking to build data centers, pushing firms to build in countries where environmental regulations and governance structures are relatively lax. While AI has the potential to help alleviate water scarcity, it is also an avaricious consumer of water, and its rollout is likely to exacerbate water inequality. One of China’s key weaknesses in the AI race is the paucity of water in the country. Its existing and future giant water management projects carry geopolitical sensitivities that could ultimately be a source of conflict with neighboring countries.
Critical minerals
AI is increasing the demand for critical minerals and raw materials, particularly for the infrastructure that support its rollout. Electric grid expansion will require copious amounts of copper, aluminum, rare earth elements, nickel, cobalt, and lithium. The expanding quantity and power of semiconductors are spurring the demand for silicon and germanium, crucial to microchips and processors. Gallium, indium, and arsenic are used in advanced semiconductors for graphics processing units and high-performance chips. Platinum group metals are used in specialized electronics and hydrogen fuel cells that may support backup systems.
Conclusion
For now, and likely years to come, the US and China will remain dependent on international trade for semiconductors and critical minerals. The drive for self-sufficiency, once achieved, could bifurcate the global economy, with competing blocs bound to one or other AI leaders as separate systems become defined by differing technology and ethical standards. AI will both enhance the volume of trade and fragment it, accentuating the already clear tendency for trade to occur within spheres of influence while broadening the variety of products that can be traded across borders.
To read the full white paper as it was posted on Hinrich Foundation: Click Here