On August 20, 2026, NL Times, citing Dutch reporting, said Washington is preparing to force the Netherlands to bar ASML from selling almost any chipmaking machines to China. The vehicle is the bipartisan Multilateral Alignment of Technology Controls on Hardware (MATCH) Act, introduced in the House on April 2, 2026 as H.R. 8170, with a Senate companion, S. 4281. As reported, it would end sales of older DUV immersion lithography tools and bar ASML from servicing machines already installed in China. Allies would have 150 days to align their export controls with US rules or face possible US sanctions. TrendForce reported on August 25 that the bill is on track to be folded into the US defense budget, and Dutch officials are lobbying against it.
The strongest case for the bill
The case for MATCH deserves a fair hearing. Export controls that bind only one supplier country leak. If US firms cannot sell a chokepoint tool to a Chinese fab but a Dutch or Japanese firm can, the restriction costs American companies revenue without slowing the target. House Foreign Affairs Chairman Brian Mast, whose committee advanced the bill on April 22, 2026, framed it as a way to "limit China's access to the most critical machines and parts needed to make advanced chips." According to the committee's release, the bill directs the Secretaries of Commerce and State to identify the chokepoints in advanced chip manufacturing that the US and its allies should jointly control. That is a coherent goal. Multilateral alignment is the only way controls work at all.
Servicing is a real lever too. An installed lithography tool without maintenance, spare parts and software support degrades over time. Bringing installed-base support into scope acknowledges that a sales ban alone leaves a large fleet running.
Where the method goes wrong
The objection here is to the mechanism, not the goal. A statutory 150-day clock, backed by the threat of sanctions and, according to reporting, an expanded foreign direct product rule reaching any tool containing US technology, treats an ally as a compliance target. The Dutch government controls ASML's export licences under its own law. Legislating that it must match US rules by a fixed date, on pain of penalty, converts a shared security project into a demand.
Three costs follow.
- Sovereignty backlash. The Netherlands has reportedly protested provisions that would override its licensing authority. Coercion pushes governments toward the very "digital sovereignty" posture that the EFF's September 9, 2026 essay describes European officials invoking in debates over semiconductors, cloud and AI. Allies who feel dictated to build independent capacity and diverge from US standards, which weakens the coordination MATCH says it wants.
- A blunt instrument for a moving target. Statutory deadlines lock in a technology definition at the moment of drafting. "Older DUV immersion tools" is a category that changes as Chinese domestic lithography advances. Licensing rules that regulators can adjust with evidence will track that better than a fixed statutory cutoff.
- Unclear evidence of the marginal gain. Restricting servicing hits mature-node production in China, much of it commercial. ASML's management said on its Q2 2026 call that China should be roughly 20% of total net sales for the full year, driven mainly by demand in mainstream logic rather than leading-edge nodes. If the aim is denying frontier AI and military chips, the bill should show why mature-node commercial fabs belong in scope.
What the numbers say
ASML reported Q2 2026 total net sales of €9.3 billion and net income of about €2.9 billion, both above guidance, with the company crediting stronger service and field option revenue. China's share of system sales has already fallen, from 19% in Q1 to 14% in Q2 according to the hook reporting. This suggests exposure is shrinking under existing licensing regimes. The market has also priced in more restriction. Analysts quoted in trade coverage put a full DUV-plus-servicing ban at a double-digit hit to revenue, and ASML's own filings list export restrictions on shipments, including of ordered systems, as a material risk to its ability to sell systems and provide services.
The servicing provision matters most here. Service revenue is stable, high-margin income that funds the R&D that keeps ASML ahead of every competitor. Cutting it may hurt the company that anchors the Western lithography lead, and the Dutch and European ecosystem around it, more than it slows Chinese fabs that can also seek workarounds and domestic substitutes.
A proportionate alternative
A better design would keep the alignment objective and change the tool.
- Negotiated, licence-based alignment. Use the Commerce and State chokepoint-identification process the bill already contains to reach agreed control lists with the Netherlands and Japan, implemented through each government's own legal authority.
- Sunset and review clauses. Tie any DUV or servicing restriction to a periodic technical review, so the definition follows the state of Chinese capability rather than 2026 drafting.
- Narrow servicing carve-ins. Restrict support for facilities tied to advanced-node or military production, and leave verified commercial mature-node fabs to case-by-case licensing.
- Consultation before sanctions. Reserve unilateral secondary measures for demonstrated non-compliance after a genuine joint process, not for a missed calendar date.
Pro-innovation policy accepts that security controls are sometimes justified. It also insists that they be narrow, evidence-based and built with allies, not imposed on them. The MATCH Act gets the first part right and risks getting the second wrong. If it rides the defense budget without amendment, Congress will have skipped the debate over scope and process that a measure this consequential warrants.
What to watch
The fights to follow are the final text of any defense-budget vehicle, whether the 150-day clock and sanctions language survive, how the Dutch government responds in its own licensing, and whether the servicing provision is narrowed. Those details, not the headline about a ban, will decide whether the outcome is durable alignment or a rift with the ally that builds the machines.