ASML CEO warns that excessive China export controls on chip-making equipment risk accelerating development of rival EUV lithography and parallel semiconductor fabs.

ASML Holding N.V. (NASDAQ: ASML) holds a critical but often overlooked monopoly in the artificial intelligence supply chain: it is the sole manufacturer of extreme ultraviolet (EUV) lithography machines required to produce advanced chips. While investors have focused on Nvidia Corp (NASDAQ: NVDA) and Taiwan Semiconductor Manufacturing Company Ltd. (NYSE: TSM), CEO Christophe Fouquet told the Financial Times that the true bottleneck lies with ASML, which supplies the equipment that makes TSMC's manufacturing possible.

Fouquet warned that overly broad export controls on semiconductor technology to China risk backfiring by accelerating Beijing's efforts to build domestic alternatives. While some restrictions are necessary for security, he argued, "over-restricting" creates desperation that incentivizes China to develop competing technologies, potentially undermining ASML's long-term dominance. Though ASML currently faces bans on selling its most advanced EUV systems to China, Fouquet highlighted that restrictions on deep ultraviolet (DUV) equipment could have unintended consequences. He pointed to Huawei's emergence as a significant force in developing DUV lithography technology as evidence that pressure can drive innovation.

The scale of ASML's technology explains why it is so difficult to replicate. An EUV machine costs roughly $400 million and is about the size of a double-decker bus, using lasers, mirrors, vacuum systems and short-wave light to print increasingly small circuits onto silicon wafers. Each machine contains more than 100,000 parts supplied by 200 principal suppliers and roughly 2,000 companies across the broader ecosystem. Companies including Zeiss, Trumpf and VDL supply critical components; ASML has acquired or invested in selected suppliers when necessary to protect the technology chain. Fouquet rejected the notion that capacity could be increased simply by injecting more capital, noting that manufacturing is constrained by the broader supply chain ecosystem.

Fouquet expects the artificial intelligence cycle itself to remain substantial, estimating the technology's effect on society and industry could last ten to fifteen years. For investors, the story extends beyond Nvidia: the AI hardware chain reaches back to a Dutch equipment maker whose technology has become one of the industry's most difficult-to-replace links.