Taiwan Deterrence Depends on Allied Industry, Logistics, and AI

Eyck Freymann sees a limited window for the United States to deter a Chinese attempt to coerce Taiwan. China holds advantages in local military mass, industrial capacity, and geography: it has more ships, aircraft, missiles, drones, and capacity to produce them, while Taiwan lies just off China’s coast and the United States would operate across an ocean. Those advantages do not make deterrence impossible. They do mean that American margins for error are shrinking.

The immediate military concern is China’s missile program. Freymann says China, never constrained by the Intermediate-Range Nuclear Forces Treaty, accumulated cruise and ballistic missiles through the 2000s and 2010s with ranges from a few hundred to a few thousand miles. Those weapons can range Taiwan and threaten targets in the Philippines, on the Korean Peninsula, in Japan, and as far as Guam—the essential logistical hub for U.S. maritime operations across the region.

A conflict could begin with an effort to saturate key bases and facilities in Japan, Guam, and elsewhere. Damaged airfields, ships, aircraft, and support infrastructure could force the United States to miss an operational step in the opening hours and days. If China then got sufficient forces onto Taiwan, toppled its government, and installed a friendly regime, it could seek to present the outcome as a fait accompli. The question would become not simply whether U.S. forces could fight, but what political and military options would remain after Taiwan’s government had fallen.

The balance is therefore not adequately captured by counting ships or aircraft. Freymann’s central premise is that deterrence is a system: the way military capabilities, industrial production, logistics, technology, alliances, and political resolve work together and adapt under pressure. A force that can survive a barrage, locate an adversary, sustain itself, replace losses, and integrate new technology can impose costs disproportionate to its numerical size.

That systems perspective informs his qualified optimism. U.S. intelligence likely has a reasonably good handle on the size of China’s order of battle—the broad number of missiles, launchers, ships, and aircraft, he says. What is harder to know is quality: missile accuracy, the performance of air and missile defenses, platform survivability, and the behavior of command, control, cyber, and space systems under wartime conditions.

War games cannot settle those questions. Freymann describes them as tools for testing assumptions and identifying ways events could go wrong, not crystal balls. The point of a simulation is to expose possible futures and force planners to confront vulnerabilities before a real crisis does. Yet he notes that Admiral Paparo has publicly said that U.S. margins of victory in simulations are decreasing. American leaders may still believe the United States could prevail if required, but the room for error is getting smaller.

The response, in Freymann’s account, is not a single procurement decision. It is a set of near-term investments: air defenses against missile salvos; submarines capable of operating close to the battlespace; resilient space systems; drones and counter-drone systems; and the logistics needed to continue fighting beyond the first week or two. These capabilities must be matched by industrial capacity and allied integration quickly enough to affect the period of greatest danger, rather than only the force structure of the late 2030s.

The Western Pacific matters to the United States, Freymann contends, not solely because of Taiwan’s political status or because a Chinese invasion would be a regional military event. The region contains high-tech supply chains essential to the modern economy: Taiwan’s semiconductor production, South Korean shipbuilding, and production of displays, controllers, industrial components, and machinery across the Indo-Pacific.

Its political geography makes these networks especially exposed. The alliance structure built in East Asia after World War II is, in Freymann’s description, hub and spoke. Unlike NATO allies in Europe, the region’s U.S. partners have not historically been as deeply connected to one another for defense; their principal security relationship has often run through Washington. The United States has treaty commitments to Japan, South Korea, Australia, and others, while Taiwan is central to the wider regional military and economic balance.

Many of these economies are also dependent on external flows. Taiwan, South Korea, Japan, and others import more than 90 percent of their energy, Freymann says, and depend on maritime and air routes for the movement of goods and people. China’s gray-zone pressure and other efforts to hold commerce at risk can therefore create leverage without an attack on the American homeland.

This is why he treats China as the United States’ “sole truly pacing challenge.” Russia’s war in Eastern Europe, Iran’s threat to Gulf partners and commercial routes, North Korea, and terrorism all impose real demands. Freymann rejects the idea that Washington can simply abandon Europe or other commitments and redirect everything to Asia; Russia, Iran, North Korea, and China are working more closely together, and such a retreat would be “penny-wise, pound-foolish.” But China combines technological capability, industrial capacity, and state capacity at a scale he considers uniquely formidable.

A Chinese ability to control the region’s commercial flows could eventually allow Beijing to coerce allied economies and use them against the United States. The issue is not merely whether China might project military power across the Pacific to attack the continental United States. It is whether it could shape the economic conditions under which America’s closest regional partners function, industrialize, and trade.

That wider exposure makes deterrence an economic-security problem. Air defense is needed in Europe, the Middle East, and the Pacific alike; demand for it is effectively unlimited, so production scale is itself a strategic constraint. Drones raise a related problem. Freymann says the United States is “shockingly dependent” on China for inputs needed to make drones domestically, including batteries, motors, and sensors. Secure supply chains with allies are therefore not separate from military readiness. They are part of the capacity to supply a coalition during a conflict.

Freymann argues that the decisive Taiwan scenario may not begin with an amphibious invasion. He agrees with Michael Boskin that China could seek to achieve its political objectives through coercive but non-kinetic means. A war ends when one side achieves its political objectives and the other accepts the result, Freymann says; if Beijing can gain control without the risks of invading and occupying Taiwan, it has reason to try.

The scenario he finds particularly concerning—and potentially the most likely—is a quarantine: a quasi-blockade in slow motion. Rather than formally sealing off Taiwan in a manner that presents an unmistakable act of war, China could pressure the companies carrying people and goods to and from the island. The objective would be to establish practical control over who and what enters and exits Taiwan while preserving ambiguity about the threshold that has been crossed.

China could seek to turn Taiwan’s semiconductor production into leverage. If it controlled the terms under which goods, people, and commercial access moved through the island without having to invade it, Beijing might tell the United States that access to Taiwanese chips could continue—but only under Chinese conditions. The pressure would be directed not only at Taiwan’s government but at the companies and foreign governments that depend on those supply chains.

That makes the contest fundamentally political. A cyberattack on Taiwan’s military systems or a propaganda campaign intended to convince Taiwanese citizens that resistance is futile might serve China’s aims. But Freymann doubts China would want to be clearly identified as the actor that shut off global access to Taiwan Semiconductor Manufacturing Company. A disruption that triggered a collapse in the Nasdaq and panic on Wall Street could also identify Beijing as the side that broke supply chains, helping its opponents assemble a broader coalition.

Public opinion is correspondingly difficult to predict. Survey responses to an abstract question about whether Americans or Taiwanese would fight cannot settle how people will react in a crisis. Freymann points to Ukraine as an example. Before Russia’s 2022 invasion, polling suggested Ukrainians did not expect to fight Russia. The failed attack on Hostomel airport, atrocities in Bucha, and Volodymyr Zelensky’s refusal to evacuate transformed public sentiment rapidly. A population that had not anticipated resistance became committed to it.

China cannot know in advance whether coercion will cause Taiwan to accept Chinese terms or trigger a rallying effect. Nor can the United States know precisely how American, Taiwanese, or allied public opinion will move. The decisions made in the opening hours and days could influence which side is regarded as responsible for interrupting global commerce, whether Taiwan’s public sees resistance as viable, and whether a wider coalition decides that action is in its own interest.

The U.S. political message, Freymann says, should be support for Taiwan’s democratic right to choose its leaders and pathway forward. The United States does not take a position on Taiwan’s ultimate status; that remains an issue for China and Taiwan to resolve. What it should reject is the loss of Taiwan’s liberty through force or coercion.

For Freymann, the stakes extend beyond continued access to chips or consumer electronics. If China can put Taiwan under sustained pressure, it could try similar methods against other regional states and compel them to structure their economies in ways hostile to U.S. economic security. He frames the confrontation as a contest over how the world’s most important economic region is organized: through allied cooperation and open commercial systems, or through a Chinese system able to control strategic economic and technological assets.

China’s ability to threaten forces it can see creates the operational problem that U.S. planning must solve. Freymann describes an anti-access/area-denial environment: a posture intended to constrain an opponent’s ability to enter and operate freely in a region. Assets on the sea surface or in the air within several hundred miles of China’s coast can be targeted. Aircraft carriers are not useless in that environment, he says, but large and relatively slow surface assets would need to operate farther from the battlespace. That in turn requires longer-range weapons, more demanding refueling arrangements, and deeper logistical support.

Attack submarines are valuable because they can operate close to the battlespace while remaining difficult to detect. U.S. attack submarines are nuclear powered, have effectively unlimited range, and can deliver substantial firepower early. Freymann calls them “the door kicker for the rest of the force”: a capability that could shape the first days or weeks of a conflict while more exposed surface forces operate under missile threat.

But the submarine question is not simply fleet size. It includes repair capacity, resupply, maintenance locations, and the pace at which aging boats are retired. If the United States cannot build enough boats to replace older vessels, Freymann warns, its fleet could shrink as China’s submarine force improves in size and quality.

Rebuilding U.S. shipbuilding is, by his account, a multidecade strategic decision that could require several tens of billions of dollars. The United States lacks not just yards and materials but skilled tradespeople, including welders and electricians. Freymann sees South Korea as particularly important because its shipbuilders know how to build vessels quickly and affordably and could help transfer practices and train U.S. workers.

He cites discussions around Hanwha Marine, a leading South Korean shipbuilder, including its interest in developing a shipyard in Philadelphia and talk of participation in Virginia-class submarine production. The sensitive propulsion technology of U.S. nuclear submarines would need to be protected, but allied firms could still contribute around those protected elements.

AUKUS, through which the United States, Britain, and Australia are working to co-produce a new submarine class, provides a longer-term path. Its timing is the problem. Freymann expects its benefits only in the 2030s, whereas he identifies the coming decade as the danger period. He suggests stopgaps that could include cooperation with South Korea on conventional, diesel-powered submarines capable of being forward deployed and maintained in the region rather than returning to the United States.

Dispersal and stockpiling address a related vulnerability. The Pacific’s distances limit the direct transfer of lessons from Ukraine. Drones there may operate a few to a few tens of kilometers behind and beyond a front line. In the Indo-Pacific, the distances are oceanic: Freymann notes that the U.S. West Coast-to-Japan distance is roughly comparable to Japan-to-Australia. Drones built for those distances cannot simply rely on batteries; they need fuel, becoming larger, louder, and bulkier.

The broader lesson nevertheless holds: a force that can be spotted is likely to be destroyed. Units must reduce their signatures, use decoys, move frequently, and avoid concentrated targets. Freymann says the Marine Corps is adapting by developing small units able to operate from islands, particularly Japan’s Ryukyu Islands north and east of Taiwan. Such units could complicate Chinese targeting, draw fire away from more valuable assets, and launch fires of their own.

Logistics is the less visible foundation of this posture. Civilian populations in the region require stockpiles to withstand disrupted trade. U.S. and allied forces need fuel, spare parts, and other essentials prepositioned in hardened and widely distributed facilities. Island allies can mitigate that risk by providing geographic access, facilities, and infrastructure to a distributed coalition.

Japan and Australia are becoming more important to this model. Freymann says Japan is doubling defense spending and investing in long-range precision munitions, ships, and other capabilities. U.S., Japanese, and Australian forces have practiced combined fires in exercises in Australia. The gain is not just additional platforms. It is the ability to combine allied geography, forces, stocks, and infrastructure into a more survivable posture.

Freymann does not treat artificial intelligence, cyber capabilities, and space systems as marginal additions to conventional military power. They are part of the connective tissue linking assets and warfighters: systems that collect and process data, communicate securely, identify targets, and determine what to strike and in what sequence.

AI can strengthen those systems, but it can also make them more vulnerable. In theory, Freymann says, anything with an on switch or off switch can be hacked through sufficiently capable electronic means. Command-and-control systems and data-processing networks are therefore both candidates for AI augmentation and potential targets for AI-enabled attack.

Maintaining an AI lead matters for deterrence and economic security alike. It also presents a governance problem that Freymann says differs from the nuclear age: some of the most strategically consequential capabilities are being developed in private laboratories. The question is not only what the government should buy. It is who should control the technology, under what terms it should be available to the U.S. government, and how a democratic state can maintain access to vital capabilities without extinguishing the innovation that produced them.

Boskin frames the tension as a problem of competing systems. The United States needs private development to continue, rather than putting technological innovation entirely under government control. But it also needs a military able to access and integrate the products of those companies for national-security purposes while competing with states in which strategic decisions are made by government direction.

Freymann points to SpaceX as an example of a commercial advantage with military implications. Its ability to recover and reuse early-stage rockets has reduced launch costs and helped it dominate proliferated low-earth-orbit satellites, he says. China does not have a SpaceX equivalent, nor, in Freymann’s assessment, an Anthropic equivalent yet. Preserving that gap requires cultivating private firms while ensuring that their activities serve the national interest as well as shareholders.

He does not offer a settled answer to the resulting questions of corporate control, government access, and democratic governance. His claim is narrower but consequential: private-sector leadership in AI and space could matter more to deterrence than marginal differences in the number of ships, planes, or drones. The United States needs the capacity to bring those technologies into a coalition faster and more effectively than China can.

Freymann’s argument creates a practical distinction between building a future force and sustaining deterrence now. Rebuilding American shipyards, expanding the submarine fleet, and bringing the AUKUS class into service are long-cycle undertakings. They may be necessary, but they do not by themselves solve the five-to-eight-year readiness interval he identifies as most dangerous.

The nearer-term task is to make coercion and rapid military action less likely to succeed: protect bases and forces against missile salvos, disperse platforms and supplies, preposition fuel and spare parts, use allied geography, and preserve advantages in undersea warfare, space, cyber, and AI. It also requires the political capacity to withstand a crisis in which China may seek an ambiguous, incremental gain rather than a plainly recognizable invasion.

Freymann’s stated aim is a peace in which both sides believe their vital interests are respected, not war with China. But he identifies Taiwan as the place where deferred conflicts of interest may be tested under Xi Jinping, whom he describes as possessing the ideological fixation, centralized military control, and industrial and technological resources to make a bid for the island. The appropriate posture, in his view, is respect without deference—and preparation sufficient to ensure that neither side mistakes uncertainty or distraction for an opportunity.