The Drone Swarm Paradox in the Taiwan Strait

The convergence of three critical timelines in 2025 reveals a dangerous paradox at the heart of cross-strait stability. Just as the U.S. Navy’s initial iteration of its “hellscape” drone swarms is expected to become operational this August, China’s jet-powered drone mothership – the Jiu Tian – is slated to begin operations this summer as well, part of the People’s Liberation Army’s (PLA) integration of unmanned systems as force multipliers for its missile-dominant strategy. At the same time, Taiwan has increasingly looked to a drone strategy, whilst also seeking to enhance its status as a drone production hub. But rather than enhancing deterrence, this technological arms race may be creating the very crisis each side hopes to avert.

China’s “Strait Thunder-2025A” exercise in April brought the Shandong carrier group within 24 nautical miles of Taiwan’s coastline, the closest major PLA exercise to the main island on record. This unprecedented proximity occurred just as both sides are racing to deploy swarm technologies, perhaps suggesting that Beijing – rather than viewing Taiwan’s current arms build-up as a purely defensive measure – interprets it as a strategic threat that must be countered before those capabilities reach full operational capacity.

Taiwan’s use of drones in its asymmetric defense strategy (the so-called “porcupine strategy”) is premised on lessons from the war in Ukraine and recent fighting in the Red Sea. However, it may fundamentally also underestimate the unique constraints of maritime warfare. In addition to creating institutional pressures within Taiwan’s military establishment to deploy drone capabilities early in any crisis, such a strategic calculus may simultaneously provide impetus for Beijing to act preemptively.

The Maritime Reality Check

Even if drone swarms have been dominating headlines from Ukraine to the Red Sea, some analysts such as Joshua Tallis have pointed out that there remains a risk of “misapplying lessons from these often land-based conflicts” without consideration for “the unique nature of war from the sea.” Although unmanned systems have had strategic effects in maritime conflicts such as Ukraine’s attacks on Russian naval vessels in the Black Sea, it would be remiss to overlook that they often worked in tandem with cruise missiles rather than as standalone platforms.

In Ukraine’s previous naval operations against the Russian Navy, about half of the major Russian platforms destroyed were lost to long-range land-attack cruise missiles targeting vessels in port. Notable operations involving drone boats, such as the October 2022 attack on the Russian Black Sea Fleet in Sevastopol, had succeeded primarily because cruise missile strikes had already degraded Russian air defenses and forced fleet consolidation. And the most significant loss – the sinking of the flagship cruiser Moskva in April – involved Ukraine’s use of two R-360 Neptune anti-ship cruise missiles rather than drones alone.

The “resiliency through mass reconstitution of cheap systems” that has defined successful land-based drone employment in Ukraine may inevitably break down across the watery expanse of the Taiwan Strait. As Tallis explained, “platform survivability and munition reconstitution are the most operationally important criteria for massing effects against the enemy” in the maritime domain. A strategy of deploying thousands of small, expendable drones ignores the fact that naval combat occurs across vast distances. “The challenges of rearming a ship, the longer tactical distances, [and] the larger necessary warheads” create crosscutting constraints that fundamentally alter the cost-effectiveness calculus for the defender.

The Missile Still Matters

While Taiwan at present has focused on its acquisition of drone swarms, it has simultaneously been pursuing more sophisticated and expensive defensive systems. In early 2025, Taipei signed a $761 million deal for U.S. NASAMS (National Advanced Surface-to-Air Missile Systems) air defense packages, and has also been actively testing air-launched variants of its indigenous Hsiung Feng-III supersonic anti-ship missiles. The air-launched Hsiung Chih variant, spotted in flight testing on the F-CK-1 Ching-Kuo fighter aircraft in February and March 2025, demonstrates Taiwan’s parallel investment in conventional precision-strike capabilities.

Conversely, Beijing has likewise continued its relentless missile buildup. Based on a 2025 Federation of American Scientists report, China now has around 710 nuclear-capable land-based missiles capable of delivering approximately 380 warheads. It was noted in mid-2024 that its nuclear arsenal had increased by roughly 100 nuclear warheads as compared with 2023. China operates five types of anti-ship ballistic missiles: the Dongfeng-21D (DF-21D), DF-26, DF-17, DF-27, as well as the hypersonic Yingji-21 (YJ-21).

Indeed, recent satellite imagery reveals a marked boost in China’s ballistic missile capabilities, particularly a substantial increase in its DF-26 intermediate-range ballistic missiles – with some reports indicating that each DF-26 brigade may even be able to launch 75 missiles at once. This missile-centric approach likely reflects Beijing's belief that the missile remains “the decisive naval munition.” In September 2024, the PLA Rocket Force would further demonstrate its operational capability through a widely publicized intercontinental ballistic missile test launch into the Pacific Ocean.

The technological gap vis-à-vis China’s military power highlights Taiwan’s strategic quandary. Despite its recent NASAMS acquisition and accelerated production of its indigenous Hsiung Feng-III supersonic anti-ship missiles – with 131 units ordered in 2022 and the Hsiung Chih cruise missile now undergoing flight testing – its conventional precision-strike capabilities remain insufficient against Beijing’s overwhelming missile arsenal.

At the same time, the kinds of advanced systems required to counter a formidable adversary such as the PLA Navy would demand greater sophistication and resilience than the relatively less high-end drones employed in the war in Ukraine. Those systems would include high data-rate satellite communications with anti-jamming capabilities and autonomous guidance systems – requirements that drive costs far above the “cheap and plentiful” threshold that makes mass reconstitution viable.

The Capability Window Paradox

Taiwan’s current defense strategy creates what might be termed a “capability window paradox.” As Taipei’s drone capabilities mature toward expected peak effectiveness by around 2025 or 2026, China faces a narrowing window of opportunity where its traditional military advantages can still decisively overwhelm swarm defenses. It also follows that this temporal compression may create incentives for Beijing to act sooner rather than later before Taiwan’s drone capabilities reach their full potential.

Moreover, other analysts also suggest that the United States’ “Replicator” program may inadvertently destabilize deterrence by creating a technological arms race that compresses decision-making timelines. Recent counter-drone advances have limited autonomy and weapon efficacy, paradoxically increasing the urgency for fully autonomous systems as militaries race to stay ahead. Such a dynamic mirrors the pre-World War I naval arms race between Britain and Germany, where technological innovation shortened the strategic decision-making timelines and reduced the opportunities for diplomatic resolution. In the present context of the Taiwan Strait, as each side develops more sophisticated autonomous capabilities, the space for human judgment and diplomatic intervention during crises may correspondingly also diminish.

Analysis of PLA flight activities in Taiwan’s Air Defense Identification Zone (ADIZ) similarly reveals escalating patterns of Chinese military pressure designed to normalize crisis-level operations. The “Strait Thunder-2025A” exercise represented a qualitative shift – not merely larger in scale than previous “Joint Sword” exercises, but operationally bolder in terms of deploying greater numbers of the PLA’s major assets past the median line of the strait, and closer to Taiwan’s contiguous zone.

Institutional momentum further compounds this temporal pressure. Once military bureaucracies have committed to specific unmanned technologies, institutional inertia and sunk costs create pressure for early deployment regardless of strategic discretion. As Jacquelyn Schneider and Julia Macdonald outlined, military organizations often adopt specific unmanned technologies based on institutional preferences, bureaucratic processes, and doctrinal assumptions – rather than purely strategic logic. Taiwan’s apparent focus on drone swarms might just create such bureaucratic momentum, even when operational conditions suggest alternative approaches may be more suitable.

Implications for Crisis Management

The proliferation of drone swarms on both sides of the Taiwan Strait introduces novel pathways to inadvertent escalation that existing crisis management frameworks are ill-equipped to address. Taipei’s predisposition toward drone swarms creates a first-mover disadvantage whereby defensive preparations become indistinguishable from offensive posturing. Indeed, the requirement to disperse and activate drone assets early in a crisis to ensure survivability could paradoxically signal hostile intent from Beijing’s perspective, compressing the period of warning available for diplomatic intervention.

This ambiguity of intent is compounded by fundamental attribution errors inherent in the use of swarm technologies. Put another way, the dual-use nature of drone concentrations – equally capable of defensive area denial or coordinated offensive strikes – creates intelligence assessment challenges that favor preemptive action over strategic patience. At the same time, China’s investments in drone swarm capabilities heightens this security dilemma, as its defense planners must assume worst-case scenarios when conceptualizing their own use of concentrated drone deployments.

The automation of response systems further erodes traditional escalation control mechanisms. As drone warfare operates at machine learning speed rather than along human decision-making timelines, the compressed reaction cycles leave insufficient space for the political oversight and strategic restraint that have historically prevented conflicts from spiraling beyond their original scope.

Taipei’s asymmetric defense strategy, while tactically sound given its resource constraints vis-à-vis its giant neighbor, may be complicated by the maritime domain of operations. Cheap, plentiful systems are more suitable for land warfare, while naval combat rewards platform survivability and sophisticated munitions – not just the mass production of attritable assets. China’s approach incorporating uncrewed platforms, which combines both drone capabilities such as the Jiu Tian as well as advanced missiles, integrates multiple systems rather than relying on single-technology solutions.

The convergence of autonomous technologies with traditional missile warfare creates new categories of escalation risk. When algorithms can trigger responses faster than human judgment can intervene, the traditional mechanisms of crisis management become inadequate. Rather than enhancing deterrence, the current trajectory toward increasingly autonomous systems may in fact be recreating the conditions for arms racing that have historically preceded major conflicts. Recognizing this paradox is a crucial first step in developing crisis management approaches that account for how rapidly evolving military technologies can outpace human oversight.