On September 24, 2026, Dutch Prime Minister Rob Jetten told Bloomberg TV he is "pretty confident we will come to some agreement that is workable for each and every one" on the US MATCH Act, according to a Stockopedia summary of the Bloomberg report. He said he has discussed the bill with President Trump and the importance of protecting the semiconductor industry. Some summaries paraphrase his wording as "very confident"; the quoted text is "pretty confident." The difference matters, because the bill's mechanics leave less room for optimism than either phrase suggests.
The strongest case for the bill
Supporters have a serious argument. Chinese fabs such as SMIC have kept advancing with deep ultraviolet (DUV) immersion tools after EUV was cut off. Unilateral US controls do little if allies keep selling and servicing equivalent machines, and a patchwork lets Beijing buy from whichever supplier is most permissive. The bill's name, Multilateral Alignment of Technology Controls on Hardware, states this logic. If the goal is to slow China's military-relevant chipmaking, aligned allied rules are more effective than rules the US applies on its own.
What the text actually does
The House version, H.R. 8170, was introduced on April 2, 2026. The Senate companion, S. 4281, was introduced on April 13 by Senators Ricketts, Kim, Risch and Schumer. It was referred to the Banking Committee. Key provisions in the House text:
- Covered equipment. It includes "deep ultraviolet immersion photolithography machines" and equipment under ECCN 3B001, 3B002 or 3B993. It also covers everything that already requires a license for export to countries of concern.
- Servicing. The bill requires "a license for all servicing of any applicable item located in any covered facility," with "a policy of denial." Servicing includes installation, calibration, repair, refurbishing and testing.
- Named entities. It lists 15 organizations, including SMIC, YMTC, Huawei and Hua Hong, for "comprehensive export controls." Early reporting described five named fabs. The text I read lists 15 entities, so any reading that stops at five understates its reach.
- The deadline. Within 150 days of enactment, agencies must either certify that allied supplier countries have adopted matching controls or list the countries that have not. A one-time 90-day extension is available if the Secretaries of Defense and Energy concur that allies are taking "concrete, verifiable steps."
The 150-day clock has not started. The bill has not been enacted, and the final text could change.
Why the servicing clause is the hard part
Banning new sales is a clean line. A servicing ban is different, because it reaches machines that Chinese fabs already own. ASML's installed-base business is a large and growing part of its revenue. In Q2 2026 total net sales were EUR 9.3 billion, and the company attributed its beat to higher installed-base management sales. That transcript also shows ASML expects China-related business to be "around 20%" of total net sales for 2026. I could not verify the reported drop in China's share of system sales from 19% to 14% against ASML's own filings, so I do not rely on it here.
A policy of denial for servicing would shut down fabs that run on tools Dutch and Japanese governments have licensed. Beijing could treat that as an act aimed at the Netherlands, not merely at Chinese firms. The bill's ultimate lever, a unilateral extension of US jurisdiction over foreign-made tools that contain US technology, would work through the Foreign Direct Product Rule. That is the kind of extraterritorial reach the European Commission has long contested in other areas.
Sovereignty without a veto
Jetten's confidence may be well placed on process. The 150-day structure invites a negotiated outcome, and the extension suggests Congress expects haggling. The problem is the framing. The Dutch government would be choosing whether to copy US rules, not setting its own. The Netherlands has already adopted national controls on advanced lithography. Those were enacted through Dutch ministerial regulation, with parliamentary oversight and its own assessment of security and economic effects. A statutory US deadline compresses that deliberation into a compliance exercise.
The pro-innovation concern is practical. Export controls work best when they are narrow, predictable and based on evidence of military end use. A blanket denial on servicing at 15 named entities, including commercial and mature-node producers, goes beyond the security rationale. It also gives ASML's customers elsewhere a reason to doubt the reliability of European suppliers, and a reason to fund domestic alternatives. China has every incentive to speed up that substitution. Controls that push rivals to build their own tools can cost the West more in the long run than they gain.
What a workable deal would look like
A deal Jetten could defend would have at least three features:
- Tiered scope. Advanced-node fabs and military end users face a denial policy. Mature-node commercial production is handled by licensing, not prohibition.
- Servicing carve-outs. Safety-critical maintenance and non-performance-enhancing parts would stay licensable. Software upgrades and capability-extending repairs would be restricted.
- Recognition of Dutch rules. The US would accept Dutch national controls as the "matching" regime, with a defined review mechanism. That would avoid the FDPR fallback while keeping Dutch parliamentary control.
None of this is guaranteed. The bill's authors chose a 150-day trigger because they want leverage. The Dutch side should accept that a deal is likely, but it should insist on shaping the technical details rather than accepting them as written.
The bottom line
Allied alignment is a legitimate goal, and the strongest argument for the MATCH Act is that fragmented controls fail. But the version in the text goes further than alignment. It mandates denial of servicing for installed tools and imposes a deadline backed by extraterritorial jurisdiction. Jetten should treat his talks with Trump as a negotiation over scope and sequencing, not as a sign that the outcome is already settled.