Beyond Compliance: Rethinking Allied Semiconductor Export Controls

Attempts at semiconductor export controls by the United States increasingly depend on technology partners such as the Netherlands, Japan, South Korea, and Taiwan to make restrictions on China effective, yet these countries do not always share the same economic and strategic interests.

In April, the House Foreign Affairs Committee advanced the bipartisan Multilateral Alignment of Technology Controls on Hardware, or MATCH Act, bringing new attention to a longstanding weakness in U.S. technology policy: American restrictions lose effectiveness when Chinese companies can obtain comparable technology from suppliers in allied countries.

But the legislation also exposes a deeper problem in the way technology alliances operate. The United States increasingly depends on countries such as the Netherlands, Japan, South Korea, and Taiwan to make semiconductor restrictions effective. Those partners share many of Washington's security concerns, but they also have their own economic interests, technological advantages, and assessments of risk.

The question is therefore no longer simply whether allies should coordinate export controls. They should. The harder question is who gets to design them.

The Netherlands, home to semiconductor-equipment manufacturer ASML, illustrates why the distinction matters.

ASML and the Dutch dilemma

ASML occupies an extraordinary position in the global semiconductor industry. Its extreme ultraviolet lithography systems are essential to manufacturing the world's most advanced chips, while its deep ultraviolet systems remain important across a much wider range of semiconductor production. ASML has been restricted from selling its EUV systems to Chinese customers since 2019. The Netherlands has progressively tightened its controls since 2023, expanding restrictions on DUV lithography equipment in 2024 and adding certain measuring and inspection technologies to its national licensing regime in April 2025.

The Netherlands shares Washington's concern that advanced semiconductor technology can present national-security risks. The Dutch government has explicitly justified its own controls on security grounds, including the potential contribution of advanced technology to military applications. At the same time, The Hague has emphasized that restrictions should be targeted and should avoid unnecessary disruption to global value chains.

That balance matters because the economic stakes are substantial. Chinese customers accounted for 29.1 percent of ASML’s total net sales in 2025, down from 36.1 percent in 2024. ASML itself notes that deliveries to China have been increasingly affected by export restrictions. Decisions over which equipment ASML can sell — and to whom — therefore affect not only allied security, but also Dutch economic interests and the strategic value of one of the country’s most important technology companies.

The MATCH Act would push this debate further. The legislation seeks greater alignment between U.S. restrictions and those imposed by allied technology-producing countries and creates a mechanism through which Washington could extend controls to certain foreign-produced items when sufficient alignment is not achieved. The Dutch government has formally raised concerns about the MATCH Act and its potential extraterritorial effects, including possible consequences for ASML's ability to service and maintain equipment already delivered to Chinese customers.

That tension does not mean the alliance is weakening. Friction like this is normal when countries with different economic stakes try to pursue the same security goals. It reflects a structural feature of the semiconductor industry: technological power is distributed among allies.

The problem goes beyond the Netherlands

ASML may be the most visible European example, but Washington faces versions of the same problem across the Indo-Pacific. Japan is a major producer of semiconductor manufacturing equipment and in 2023 added 23 types of advanced semiconductor-manufacturing equipment to its export-control regime. Tokyo emphasized that the measure was intended to prevent military diversion and maintain international peace and security, while also noting that Japan had consulted with allies and partners, including the United States and the Netherlands.

South Korea faces a different calculation. Samsung and SK Hynix maintain substantial semiconductor manufacturing capacity in China; according to Taiwan's Ministry of Economic Affairs, roughly 20 percent of Samsung's DRAM capacity and 40 percent of SK Hynix's is located there. Taiwan occupies yet another position. It accounts for more than 90 percent of global leading-edge chip manufacturing and has itself tightened technology controls, adding Huawei and SMIC to its Strategic High-Tech Commodities Export Control Entity List in 2025. At the same time, Taiwan faces a uniquely direct security challenge from Beijing. These countries share many interests, yet their positions begin to diverge once economic risks or domestic industrial priorities come into play.

That is precisely why a system based primarily on allies matching restrictions designed in Washington is likely to encounter resistance. The countries whose cooperation the United States needs are not simply markets in which American policy must be implemented. They control technologies, companies, production capacity, and expertise without which the policy may not work.

The real challenge is to use that technological interdependence as something that supports cooperation instead of letting it become a constant source of tension between allies.

The limits of compliance

The logic behind the MATCH Act is understandable. Diplomacy without consequences may not produce sufficient alignment. If an American manufacturer is prohibited from supplying technology to a Chinese company while a Dutch or Japanese competitor can provide an equivalent product, restrictions can shift trade rather than constrain the targeted capability.

There is evidence that this can happen. Research by the European Commission's Joint Research Centre found that U.S. semiconductor export controls reduced targeted U.S. exports to China, while exports of semiconductor manufacturing equipment from the EU, Japan, and Singapore increased significantly.

Coordination is clearly needed, yet it is not the same as compliance, and treating them as identical often obscures the real policy differences between allies. Compliance asks whether the Netherlands, Japan, South Korea, or Taiwan has moved sufficiently close to a policy developed primarily in Washington. Co-design begins earlier. It asks the allies together which Chinese capabilities pose an unacceptable security risk, which technologies actually enable those capabilities, and which controls are most likely to constrain them without imposing unnecessary economic and diplomatic costs.

As the debate shifts toward less advanced technologies, that boundary starts to matter much more, and many of the new measures are moving in that direction. The security argument for denying technologies directly enabling cutting-edge military, artificial-intelligence, or advanced-computing capabilities can be strong. But the calculation becomes more complicated as controls extend toward less advanced equipment whose contribution to China's technological development is more indirect.

There comes a point when allies need to separate focused national‑security measures from a wider effort to contain China’s technological development, because the two approaches carry very different implications. If that boundary becomes unclear, governments risk weakening domestic and allied support for the restrictions that matter most.

From compliance to co-design

A stronger allied system should begin with three principles. First, define capabilities before products. Washington, The Hague, Tokyo, Seoul, Taipei, and other relevant partners should jointly identify the military and advanced-computing capabilities they intend to constrain before negotiating lists of controlled equipment. This reverses the usual political debate. Instead of beginning with the question of whether an ally has adopted a sufficiently similar product list, governments would begin with the security outcome they collectively want to achieve. That distinction matters because technology evolves faster than export-control lists. A capability-based approach would allow governments to ask whether a particular machine, component, software package, or service materially enables the capability they are trying to restrict rather than treating the expansion of product lists as an objective in itself.

Second, incorporate allied costs into the design of controls. Economic costs do not automatically outweigh national-security concerns. But neither should they be treated as irrelevant. If a strategy depends on Dutch, Japanese, Korean, or Taiwanese companies surrendering access to important markets, their governments should have meaningful influence over where restrictions are drawn. Economic interests are already part of the American debate. Supporters of the MATCH Act argue that uneven controls put American semiconductor-equipment companies at a competitive disadvantage relative to foreign suppliers. That concern may be legitimate. But it also reinforces the case for co-design. Measures intended to protect collective security should not simultaneously be perceived by allies as reallocating commercial advantage among them.

Third, evaluate controls for effectiveness. What matters in the end is not how long or detailed the control list becomes, but whether the measures actually slow the Chinese capabilities governments want to constrain. Governments should periodically assess whether each restriction has actually slowed the targeted Chinese capability, whether trade has shifted to alternative suppliers, whether indigenous Chinese alternatives have emerged, whether enforcement remains practical, and whether the security benefit continues to outweigh the economic and diplomatic costs. If governments see these measures as joint security tools, it makes sense for them to assess the impact together and to identify where the controls fall short.

What co-design could look like

None of this necessarily requires another large international organization. The major semiconductor technology economies could instead establish a standing coordination mechanism focused specifically on advanced semiconductor capabilities. The United States, the Netherlands, Japan, South Korea, and Taiwan would be natural participants, with other countries included when their technologies become relevant. Its most important function would occur before new controls are announced.

Participating governments could conduct joint assessments of the capabilities they seek to restrict, identify the technological chokepoints that genuinely matter, examine the economic exposure of each partner, and agree in advance on how effectiveness will be measured.

The purpose would not be to give every country a veto. National governments would retain authority over their own export-control regimes. The purpose would be to ensure that policies presented as collective security measures are collectively designed before governments are pressured to align after the fact.

Such cooperation would also make controls more durable. A restriction imposed reluctantly under external pressure is politically easier to reverse than one a government helped design and can defend as serving its own security interests.

Allied ownership

The Netherlands should not seek equidistance between Washington and Beijing. It is a NATO ally of the United States, and Dutch policy already recognizes that advanced semiconductor technology can have genuine national-security implications. But alliance does not require identical interests.

The semiconductor supply chain demonstrates why. No government independently controls all the lithography equipment, deposition and etching tools, design software, materials, intellectual property, fabrication capacity, and advanced packaging required to produce the world's most sophisticated chips.

That interdependence is often described as a vulnerability. It can also be a strategic advantage.

Washington's ability to restrict China's access to advanced semiconductor technology is much greater when the countries controlling key chokepoints act together. But that leverage is sustainable only if those countries believe they have meaningful ownership of the policies governing their technologies.

The MATCH Act addresses a genuine problem: asymmetric export controls. Where it risks falling short is in treating alignment as an outcome to be obtained from allies rather than a policy to be constructed with them. Without that ownership, Washington risks creating controls that become increasingly expansive but progressively harder to sustain as a common allied strategy.

For Washington, the lesson from ASML should therefore extend well beyond the Netherlands: effective technology alliances require more than allied compliance. They require allied ownership of the strategy.

DISCLAIMER: All views expressed are those of the writer and do not necessarily represent that of IIPA and this platform.

---

Author

Hannah Zhang is a Dutch student at The Taft School in Connecticut. Her research interests include international relations, U.S.-China competition, and the strategic choices facing smaller states. Her previous work has been published by Pacific Forum, where she examined China’s growing influence in Colombia through manufacturing, engineering, infrastructure, and integrated supply chains.

---