Silicon shield

The silicon shield is the contested thesis that Taiwan's dominance of advanced semiconductor manufacturing deters a Chinese attack, because the destruction or seizure of the island's foundries would impose intolerable costs on China, the United States, and the world economy alike. The claim is double-sided: the shield is held both to deter Beijing from invasion and to bind Washington to Taiwan's defence.

Origin

The phrase was coined by the technology journalist Craig Addison in a September 2000 International Herald Tribune essay and developed in his book Silicon Shield: Taiwan's Protection Against Chinese Attack (2001), which argued that the information economy's dependence on Taiwanese chipmaking had replaced US carrier groups as the island's real security guarantee. Current concentration must be sourced by product, process node, time and denominator. TSMC's 2025 annual report documents its technologies, Taiwanese facilities and overseas expansion; it does not itself measure deterrence or provide a universal share of "advanced chips".

Mechanism and refinements

The deterrent logic is economic: invasion or blockade could interrupt foundry output and wider supply chains, creating costs for China, the United States and other economies (Taiwan blockade scenario (economic dimension)). The magnitude and decision effect remain model-dependent, so the inherited ten-per-cent-of-world-GDP figure is removed. Recent arguments relocate the shield from one foundry to the network of non-substitutable inputs, including Extreme ultraviolet (EUV) lithography, optics, design software and chemicals. That dependence may make seized facilities hard to operate, but it does not establish that Beijing expects seizure, destruction or continued operation, and this entry does not rely on an unverified remote-disable claim.

Contestation

The thesis is contested at every joint. Rationality: economic cost may not dominate political and military objectives, so dependence cannot stand in for evidence of decision-maker belief. Perverse incentive: some analysts argue concentration could invite seizure or coercion rather than deter it. McKinney and Harris's "broken nest" proposal, under which Taiwan would threaten denial of semiconductor facilities, is a different scorched-earth deterrent rather than evidence that the passive shield works. Self-erosion: overseas fabrication and supply-chain diversification may reduce geographic concentration, but firm investment can reflect resilience or commercial motives rather than a coordinated state effort to weaken the shield. The thesis therefore remains a contested deterrence hypothesis.

Government policy, firm geography and adversary belief are separate variables. TSMC can diversify capacity for resilience or commercial reasons; Washington can support overseas fabrication; and Beijing can assess military and political objectives under a different risk calculus. None proves that the shield works. A deterrence test therefore needs evidence of capability, communicated consequence, decision-maker perception and restraint, while preserving rival explanations for non-attack.

The unit of analysis also matters. Leading-edge fabrication, mature-node capacity, packaging, equipment and design software have different geographies and substitution times. A concentration claim must specify the product, process node, date and denominator before it can support a deterrence argument.

See also

Taiwan · TSMC · Chokepoint deterrence (chain of peace) · Taiwan blockade scenario (economic dimension) · Semiconductor chokepoint · Semiconductor foundry capacity (TSMC, Samsung, GlobalFoundries) · Extreme ultraviolet (EUV) lithography · Weaponised interdependence · Economic deterrence · Economic statecraft

Sources

• TSMC, 2025 annual report.
• TSMC, full 2025 annual report.
• Craig Addison, *Silicon Shield: Taiwan's Protection Against Chinese Attack*, bibliographic record.
• Jared M. McKinney and Peter Harris, "Broken Nest", *Parameters* 51, no. 4 (2021).