Chinese restrictions on tungsten threaten AI chip production chain

1–2 minutes

New export rules introduced by China in February 2025 are putting pressure on the supply of tungsten micro-drills, essential tools for manufacturing circuit boards and chip substrates used in artificial intelligence systems.

China’s share of the global tungsten market

According to the United States Geological Survey (USGS), China extracted approximately 67,000 tons of tungsten in 2025, accounting for around 79% of global production, estimated at 85,000 tons. This concentration means any changes to the country’s export rules have significant implications for manufacturers worldwide.

How the new controls work

The measures announced by China’s Ministry of Commerce apply not only to tungsten but also to tellurium, bismuth, molybdenum, and indium. Companies now need special licenses to export these materials. While the controls do not amount to a total ban, they require prior authorization.

Impact on production tools

Japanese company Union Tool provides a good example of this process: its drills are made from tungsten carbide bars, which undergo cutting, assembly, and rigorous inspections. These tools are used in electronic boards, forming part of the supply chain for AI equipment.

The reliance on tungsten is less visible than that on lithography machines but is equally critical. Every stage of chip production depends on specific inputs, tools, and components, and a shortage of any one of these can halt the entire process.

Alternatives and their limitations

The USGS notes that materials such as ceramics and other carbides can replace tungsten in some applications. However, many of these alternatives either reduce the amount of metal used without eliminating it entirely or increase costs and compromise tool performance.

So far, there is no data quantifying losses in chip production or delays in data centers directly linked to the restrictions. The risk to the electronics supply chain lies in the geographical concentration of an essential input and the lack of substitutes that offer the same properties.

In summary, China’s restrictions highlight how the expansion of artificial intelligence is intrinsically linked to mineral resources and industrial tools, creating a vulnerability that could affect the costs and delivery times of AI equipment in the future.