Executive Summary
Western export-control policy treats extreme ultraviolet lithography (EUV), embodied in the Dutch lithography manufacturer ASML, as a durable barrier to China’s most advanced chips. It is neither durable nor ASML’s alone. The chokepoint is a coalition held across Dutch, German, Japanese, and American suppliers, and China is moving to bypass rather than rebuild it, placing a measurable expiration date on the embargo. Policymakers should rebalance accordingly, spending less on defending a depreciating barrier and more on the advantages that compound. That means targeted allied investment in the frontier’s true bottleneck, controls aimed at China’s alternative programs, and coordinated protection of the wider ecosystem.
Introduction & Background
Extreme ultraviolet lithography (EUV) is the foundation of modern computing power. The most advanced chips in the world, the predecessors behind artificial intelligence, modern telecommunications, and precision weapons systems, cannot be manufactured without it. An EUV machine is among the most complex devices ever engineered, and at present, only one company, the Dutch firm ASML, can build one. (Potkin 2025) This monopoly has made EUV a central chokepoint in the technological competition between China and the West, and the foundation of a decade of Western export-control policy designed to deny Beijing access to it. (Nikol-Schwarz 2026)
The durability of that chokepoint is now in question as officials in the United States informed ASML that they had evidence one of its restricted systems had reached China, a claim the company denies and that remains unresolved. (Shah et al. 2026) Whatever its resolution, the episode highlights a deeper uncertainty: policymakers understand far less than they assume about where Western leverage truly lies and how long it can be expected to hold.
The conventional account is straightforward as the leverage belongs to ASML and the task to keep its machines out of the hands of China. (Stroobants 2025) This is both incomplete and misleading, misidentifying where the leverage actually resides and misjudging how long it will last. The prevailing debate compounds the error, oscillating between alarm that China has already mastered EUV technology (Pao 2025) and dismissal that it remains almost a decade behind the West. (The Economist 2026) Neither position withstands scrutiny, and a policy built on either invites strategic miscalculation. The leverage the West holds is real, but it is narrower and more fragile than conventional account allows. Understanding its true shape is the precondition for any strategy that hopes to extend it.
Problem
The assumption that Western leverage begins and ends with ASML obscures its true nature: the chokepoint is not a company but instead, a coalition. The capability to build an EUV machine is distributed across a small set of allied jurisdictions, each holding an irreplicable piece. The Netherlands, through ASML, integrates the system. Germany supplies what no one else can manufacture, the optics from Carl Zeiss SMT, whose production capacity sets the tempo for the entire industry, and the drive lasers from TRUMPF. (Tan 2026) Japan provides the specialized chemicals, as roughly 95% of high-end photoresist are processed by three firms: JSR, Tokyo Ohka Kogyo, and Shin-Etsu. (Song 2025) The United States contributes foundational components and, increasingly, political pressure that binds the others to a common line. (Swanson et al. 2024) While Western leverage is real, it is collective and holds only as long as the coalition of firms and states continue to act in collaboration.
China, on the other hand, is pursuing a different strategy, opting for domestic production rather than a supply chain spread across multiple allied states (Tan 2026. While bypassing the West’s greatest vulnerability, China is also generating EUV light through fundamentally different physics so that the coalition’s chokepoints cease to matter. Rather than a single program, this consists of multiple companies advancing towards a common objective. Shanghai Micro Electronics Equipment builds conventional deep ultraviolet tools, a domestic workhorse rather than a path forward. (Wu et al. 2025) Additionally, a consortium centered on Huawei and Shenzehen SiCarrier has validated a prototype using an alternative light source that does not yet reach production-grade power but is closing the gap. (Pan et al. 2025) And a longer-range national program at Tsinghua University pursues a particle-accelerator source that, if realized, would render the Western monopoly obsolete rather than merely circumvent it. (Tan 2026) The distinction the public debate misses is that these are attempts not to catch up, but to change the terrain.
The strategic consequence is that the embargo carries an expiration date. Export controls are constraining China now, and their access to Western tools has limited quickly, with its share of ASML’s system sales expected to fall from roughly 33 to 20% this year. (Sterling 2026) However, denial buys time, not permanence, and the clock is set by the maturity of Beijing’s alternatives rather than by the stringency of the controls themselves. A policy that treats the chokepoint as durable is planning for a world that is already passing.
Even this understates the challenge, as control of the machine was never the true measure of the advantage. Possession of an EUV tool would place China at the beginning of the problem, not its resolution. Translating capability into competitive production requires specialized materials, accumulated industrial knowledge, and years of refinement, all of which are strengths the Western coalition holds across an entire ecosystem, not in a single instrument. Current strategy fixates on the one asset that is depreciating while neglecting the many that are not.
Possible Solutions
If the chokepoint is collective and its expiration date already set, the task is no longer to preserve Western leverage but to redeploy it. The first approach is to tighten the blockade by extending controls from EUV to older deep-ultraviolet tools, restrict the servicing of machines already installed in China, and compel allied capitals to align on pain of unilateral action, as proposed legislation in Washington now contemplates. (Dalton 2026) This is the instinctive response that also buys time. But it is defense of a depreciating position: controls leak at the point of enforcement rather than design, as the unresolved dispute over a possibly diverted machine illustrates, and every increment of tightening purchases a shrinking interval before Beijing’s alternatives mature. (Alessandria et al. 2024)
The second approach is to extend the frontier itself, accelerating the next generations of lithography on the logic that a moving lead cannot be closed by an adversary aiming at where it used to be. Its appeal is that it competes on capability rather than denial. However, the approach carries a few liabilities. It is enormously expensive and slow, with payoffs measured in years and no guarantee of return; and the pace of advance runs through a handful of irreplaceable suppliers, especially pertaining to Carl Zeiss, whose capacity to manufacture ever more demanding optics sets the tempo for each new generation of tools. (Hoeijmakers 2026) Generic calls to “invest more” accomplish little unless aimed precisely at those constraints. (Tan 2026) More fundamentally, staying ahead at the frontier does not foreclose China’s path if Beijing is pursuing a different route entirely: a novel light source that leapfrogs the current trajectory would not be caught by extending it. Frontier leadership secures the West’s position only if that contest is actually decided at the frontier, an assumption worth interrogating rather than presuming.
The offense- oriented approaches face an obstacle rarely acknowledged in strategic discussion: the structure of Western capital works against them. The alliance’s export controls are themselves the largest drag on market value of the very firms it relies upon, and deep entanglement with government on the form of subsidies, guaranteed purchases, and equity, tend to depress a company’s valuation and invite political risk, giving firms reason to resist the partnerships the strategy requires. (Atkinson 2026) China’s state-directed champions face no such discipline. For long-term work with no near-term commercial return, a market-financed ecosystem is structurally disadvantaged against a state-financed one. This is certainly an asymmetry the West must offset without abandoning the market model that remains its greater strength.
These approaches are complementary rather than competing. Denial buys the time that investment requires to bear fruit, and the two approaches will likely either succeed together or not at all. The deficiency in the current policy lies not in the choice between them, but in the proportion.
Recommendations
The challenge is not whether to maintain denial, but how to rebalance its role by redirecting resources toward sources of leverage that generate cumulative advantages.
First, the European Union and the German government should make Zeiss’ optics capacity an explicit object of industrial policy. Rather than spread subsidies diffusely across the semiconductor sector, the United States and its European allies should direct targeted co-investment and long-term capacity commitments at specific optics-manufacturing constraint that gates the frontier, treating it with the strategic seriousness now reserved for fabrication capacity. (Swanson et al. 2024) Policymakers should establish a dedicated allied compact around this single bottleneck rather than fund another broad chips subsidy.
Second, the United States and allied governments should deploy public capital through instruments that carry the long-horizon risk industry will not. Governments should use sovereign co-investment vehicles, guaranteed offtake agreements, and milestone-based grants, modeled on defense procurement and advance-purchase mechanisms rather than open-ended research funding to underwrite next-generation lithography without forcing firms to absorb losses their shareholders would punish. (Hufbauer 2025) Structuring support as a national-security investment, not a corporate handout, is what will make this approach more politically durable and financially acceptable to firms.
Third, Export-control authorities in the United States, Netherlands, Germany, and Japan, should shift the regime from finished systems toward the upstream inputs of China’s alternative programs. Allied governments should update their control lists to cover the components and materials that laser-discharge and accelerator-based light sources depend on, a category in which Germany’s laser and photonics industry, led by TRUMPF, is a decisive supplier. (Allen 2024) Enforcement, the persistent weak point, should be tightened through coordinated end-use verification and closure of servicing and secondhand-tool channels. A regime aimed where China is investing is worth more than one aimed at machines it no longer receives. (Chan 2026)
Fourth, allied governments should treat the surrounding ecosystem as a strategic asset. They should protect their advantages in photoresists, design software, and process expertise through a standing allied mechanism rather than ad hoc national measures the consensus-bound Wassenaar Arrangement, to which China’s own suppliers are not even party, and pair it with a policy that retains critical expertise while restricting its transfer. These are the assets no single breakthrough can neutralize. (Bela 2026)
Works Cited
Alessandria, G.A., Khan, S.Y., Khederlarian, A., Ruhl, K.J. and Steinberg, J.B. (2024) Trade War and Peace: U.S.-China Trade and Tariff Risk from 2015–2050. NBER Working Paper 32150. Cambridge, MA: National Bureau of Economic Research. Available at: https://www.nber.org/system/files/working_papers/w32150/w32150.pdf
Allen, G.C. (2024) The True Impact of Allied Export Controls on the U.S. and Chinese Semiconductor Manufacturing Equipment Industries. Washington, DC: Center for Strategic and International Studies, 26 November. Available at: https://www.csis.org/analysis/true-impact-allied-export-controls-us-and-chinese-semiconductor-manufacturing-equipment
Atkinson, R.D. (2026) Mobilizing for Techno-Economic War, Part 3: Transforming Financial Capitalism Into National Power Capitalism. Washington, DC: Information Technology and Innovation Foundation, 13 April. Available at: https://itif.org/publications/2026/04/13/mobilizing-for-techno-economic-war-part-3-national-power-capitalism/
Bela, V. (2026) ‘Wassenaar Arrangement: will China join or kill this 30-year-old weapon control club?’, South China Morning Post, 24 February. Available at: https://www.scmp.com/news/china/science/article/3342608/wassenaar-arrangement-will-china-join-or-kill-30-year-old-weapon-control-club
Chan, K. (2026) ‘Competing AI strategies for the US and China’, Brookings Institution, 16 April. Available at: https://www.brookings.edu/articles/competing-ai-strategies-for-the-us-and-china/
Dalton, M. (2026) ‘The MATCH Act is industrial policy dressed as national security’, R Street Institute, 30 April. Available at: https://www.rstreet.org/commentary/the-match-act-is-industrial-policy-dressed-as-national-security/
Hoeijmakers, R. (2025) ‘How ZEISS and ASML enable the modern chip industry’, hoeijmakers.net, 19 March. Available at: https://hoeijmakers.net/zeiss-and-asml/
Hufbauer, G.C. and Hogan, M. (2025) Industrial Policy Through the CHIPS and Science Act: A Preliminary Report. PIIE Briefing 25-1. Washington, DC: Peterson Institute for International Economics, January. Available at: https://www.piie.com/publications/piie-briefings/2025/industrial-policy-through-chips-and-science-act-preliminary-report
Nicol-Schwarz, K. (2026) ‘U.S.-China AI feud sees ASML walk tightrope between sales and geopolitics’, CNBC, 17 July. Available at: https://www.cnbc.com/2026/07/17/us-china-ai-feud-asml-tightrope-sales-geopolitics.html
Pan, C. and Goh, B. (2025) ‘Factbox: Key products in Huawei’s AI chips and computing power roadmap’, Reuters, 18 September. Available at: https://www.reuters.com/world/china/key-products-huaweis-ai-chips-computing-power-roadmap-2025-09-18/
Pao, J. (2025) ‘Made-in-China EUV machine targets AI chip output by 2028’, Asia Times, 19 December. Available at: https://asiatimes.com/2025/12/made-in-china-euv-machine-targets-ai-chip-output-by-2028/
Potkin, F. and Alper, A. (2025) ‘How China built its “Manhattan Project” in AI chips’, Reuters, 17 December. Available at: https://www.reuters.com/world/china/how-china-built-its-manhattan-project-rival-west-ai-chips-2025-12-17/
Seoul Economic Daily (2025) ‘Even rumors of “suspension of photoresist exports”… the conflict between China and Japan grows’, 1 December. Available at: https://en.sedaily.com/international/2025/12/01/even-rumors-of-suspension-of-photoresist-exports-the
Shah, A., Dey, M. and Peverieri, A. (2026) ‘US tells ASML it is concerned China may have top chip tool’, Reuters, 19 June. Available at: https://www.reuters.com/world/china/us-tells-asml-it-is-concerned-china-may-have-top-chip-tool-bloomberg-news-2026-06-19/
Sterling, T. (2026) ‘US targets Chinese chipmaking with proposed export restrictions on ASML and others’, Reuters, 3 April. Available at: https://www.reuters.com/world/china/us-targets-chinese-chipmaking-with-proposed-export-restrictions-asml-others-2026-04-03/
Stroobants, J.-P. (2025) ‘Semiconductors: ASML, the tech giant at the center of the US-China conflict in the race for AI’, Le Monde, 26 December. Available at: https://www.lemonde.fr/en/economy/article/2025/12/26/semiconductors-asml-the-tech-giant-at-the-center-of-the-us-china-conflict-in-the-race-for-ai_6748844_19.html
Swanson, A. and Gamio, L. (2024) ‘The global effort to make an American microchip’, The New York Times, 20 March. Available at: https://www.nytimes.com/interactive/2024/03/20/business/economy/semiconductor-chip-manufacturing-us.html
Tan, N. (2026) ‘China’s EUV lithography progress: parsing signal from noise’, The Diplomat, July. Available at: https://thediplomat.com/2026/07/chinas-euv-lithography-progress-parsing-signal-from-noise/
Wu, Z. and Olcott, E. (2025) ‘China trials its first advanced tools for AI chipmaking’, Financial Times, 16 September. Available at: https://www.ft.com/content/8fd79522-e34f-4633-bc87-ef0aae2d9159