The Innovation Monopoly: Why ASML and TSMC Break Antitrust

Introduction

Semiconductors are the most important element in modern age’s infrastructure. They are often called the “oil of the 21^st century” as they are used in almost everything around us, whether it’s our smartphones, laptops, cars, or even in military equipment.

Two giants sit at the absolute apex of this significant supply chain. ASML, a Dutch company that manufactures the extreme ultraviolet lithography machines that are necessary to produce the world’s most advanced chips. It holds an unbelievable 100 percent market share in that category. TSMC, Taiwan’s pride, is the second part of this supply chain, with a market share of 72 per cent. It makes 90 per cent of the advanced chips.

When two companies control such technologies without which advanced computation is simply impossible, competition law faces a question it was not designed to answer before.

Understanding The Monopoly

To comprehend the legal question, one must first understand what is being monopolised. Semiconductor fabrication is done via nodes (historically referred to the size of the smallest feature of a transistor), measured in nanometres, which represents the size of the transistors printed on a chip.

To achieve this miraculous precision, something called Extreme Ultraviolet (EUV) lithography is used. As EUV light doesn’t exist naturally, it must be generated artificially via the machines made by ASML, which cost $250 million each. This machine uses laser pulses to vaporise droplets of molten tin at a rate of 50,000 times per second. The resulting plasma emits extreme ultraviolet light, which has a wavelength of just 13.5 nanometres, short enough to print incredibly dense patterns onto silicon wafers.

TSMC, using its access to these machines and 30 years of chip manufacturing experience, has gained a hold in this industry which no other company is able to compete with. Its competitors, Samsung and Intel, both manufacture these advanced chips, but neither has been able to achieve the same yield rates and challenge the monopoly created by TSMC. This isn’t a case of traditional market monopoly but rather a technical one.

Why Traditional Antitrust Falls Short: The Innovation Monopoly Paradox

ASML and TSMC don’t fit in the traditional antitrust doctrine, it rests on a straightforward assumption: power concentrated in the hands of one company harms consumers through higher prices and negligible innovation. The Sherman Act, EU’s Article 102 TFEU, and their equivalents globally are based on this very assumption.

First, their dominance came from dedication and innovation, not exclusionary conduct. ASML invests 17 per cent of its annual revenue in R&D, which helped it to outcompete Nikon and Canon not by excluding them but by building something which they could not replicate.

Second, pricing is not always the issue. TSMC’s advanced node pricing is competitive, and ASML sells its machines to any qualified buyer who can afford them and is not subject to any export restrictions (except to China). There’s no predatory pricing or cartelization involved, just a natural monopoly born of cumulative innovation, which is treated with caution by the contemporary antitrust frameworks. This isn’t a case of traditional monopoly but rather a structural one, as if this one firm falls, the whole society would have to bear the consequences of it. The traditional antitrust doctrine isn’t even designed to label or handle that kind of risk.

Export Controls and the WTO Fault Line

While competition law is busy debating, governments are already rewriting the rules via export restrictions, most aggressively by the US. The CHIPS and Science Act of 2022 allocated 52 billion USD for domestic manufacturing, while restricting recipients from expanding production in China for ten years.

The US government has pressured the Dutch government to restrict ASML from exporting its highly advanced EUV machines to China. In return, China, has responded with export restrictions on rare earths, which ASML needs for its machines. Due to this, we are now stuck in a tech war, which the current trade laws weren’t built to handle.

Article XI of GATT generally prohibits export restrictions, while Article XXI provides a national security loophole to restrict the export of goods. The 2019 Russia-Traffic in Transit ruling suggested that the WTO would not just let countries lie about security concerns, but they are still terrified of telling a sovereign nation what is or isn’t a threat to their security. This resulted in a framework that lets countries weaponise supply chains, hold essential goods hostage, and leave no neutral party left to decide when a country is actually “protecting itself” versus when it’s just being a bully.

The Essential Facility Doctrine: Promising but Incomplete

Essential facility doctrine, which traces back to  United States v. Terminal Railroad Ass’n (1912), and developed further through cases like  B&I Line v. Sealink Harbours (1992), holds that a dominant firm which is in control of an important infrastructure must share it with its competitors and not hold onto it.

EUV machines are irreplaceable in advanced chip manufacturing, and no other company builds them. But the analogy breaks down fast. ASML nearly went bankrupt several times. Chipmakers like Intel, Samsung, and TSMC had to fund the EUV development. That’s not a monopoly that was captured, it was one that barely survived into existence.

EU courts require that a refusal to deal must shut out competition. However, ASML doesn’t refuse to sell, rather it can’t even fill its existing orders. The export restriction argument is controversial as the Dutch government, under sustained US pressure, revoked mid-shipment licenses and framed it as a “national security” call. But when a government can reach into a private company’s order book and treat it as a foreign policy lever, the usual protections around IP and contract start looking a lot weaker than they’re supposed to be.

Reforming the Framework: What Law Should Actually Do

The gap between the problem and the doctrine calls for creativity rather than doctrinal stretching. Three reforms deserve serious attention.

1. FRAND Licensing Frameworks- The EU Chips Act should legally compel access to critical technologies on fair, regulated terms, not free access, but a licensed access modelled on telecommunications standard-essential patents. Qualcomm’s “No License, No Chips” policy showed what unchecked chokepoint leverage looks like. Semiconductor equipment poses the same risk on a larger scale.

2. Transparency Over Forced Transfer- Forced licensing would cut innovation incentives. Instead, dominant suppliers should disclose technical information to governmental bodies, so governments can make smarter policy decisions without exposing trade secrets.

3. A Multilateral Dual-Use Framework- The Wassenaar Arrangement is breaking down as Russia alone can veto new controls. China held 36 per cent of ASML’s revenue in Q4 2025 through DUV stockpiling, a direct symptom of cliff-edge rules rather than predictable ones. Create a neutral forum where countries justify “national security” export bans. As right now, they can weaponize supply chains with zero accountability.

Conclusion

ASML and TSMC enjoy a position that traditional competition law wasn’t built for. They are not monopolists by law, but their control over essential technologies creates a stability risk that current frameworks simply ignore. For this, we need new regulatory frameworks that distinguish between monopolies that harm consumers and those that threaten systematic infrastructure. Otherwise, doctrine will fall behind a rapidly changing supply chain.

---

Authored by Saransh Patwal, first year student at Rajiv Gandhi National University of Law, Punjab.