Toplines

Taiwan’s National Chung-Shan Institute of Science and Technology (NCSIST) revealed details of a hypersonic missile interceptor that could be central to the proposed T-Dome air defense concept. Taiwan’s legislative gridlock has delayed and obstructed Taiwan’s development of advanced air defenses. NCSIST, a state-owned corporation that serves as Taiwan’s primary military research and development hub, disclosed the specifications of the Chiang Kung (Strong Bow) interceptor on August 22.[1] NCSIST said the Chiang Kung, also known as the Tien Kung IV (Sky Bow IV), would use a land-based mobile vertical launch system to fire interceptor missiles at up to hypersonic speeds to intercept ballistic missiles at “mid-level” altitudes — up to 70 km, according to an unnamed defense official, which is a major improvement on the Tien Kung III missile’s maximum altitude of 45 km.[2] Each launcher can hold four missiles at a time. The missile will use a two-stage solid-fuel rocket to propel the warhead and thrust-vector control to maneuver it at high altitudes.[3] The Chiang Kung system includes a mobile active electronically scanned array radar vehicle co-developed by Taiwan and the United States, which can classify, track, and engage multiple tactical ballistic missiles while being resistant to electronic jamming.[4]

The Chiang Kung will form the outermost layer of the multilayered T-Dome integrated air and missile defense (IAMD) concept, which Taiwanese President William Lai Ching-te proposed in October 2025.[5] Taiwan’s Ministry of National Defense (MND) plans to build two Chiang Kung systems with at least 128 missiles on a proposed budget of $1.13 billion.[6] The T-Dome will also include Tien Kung III and Patriot Advanced Capability-3 (PAC-3) missiles.[7] The Chiang Kung could improve Taiwan’s interception capacity and speed and reduce Taiwan’s vulnerability to the People’s Liberation Army (PLA) missile strikes. Indigenous production of the Chiang Kung system will reduce Taiwan’s dependence on procurements and technical assistance from the United States.

Taiwan’s legislative gridlock remains a major obstacle to Taiwan actualizing the T-Dome concept. Taiwan’s Cabinet initially proposed funding for Chiang Kung missile production as part of its 2026 special defense budget, but the final budget bill that passed in May excluded funding for the missile and other domestically produced systems.[8]

Taiwan’s Keelung District Prosecutors Office announced indictments for nine individuals, which included one Nvidia employee and two Super Micro employees, for smuggling Super Micro advanced artificial intelligence (AI) servers containing high-end Nvidia chips to the People’s Republic of China (PRC) in violation of US export control regulations.[9] Prosecutors alleged that the defendants attempted to claim that the 130 B300 Super Micro servers were for a rented server facility in Taiwan, even though the proposed facility lacked sufficient rack space, power, and bandwidth to use the servers.[10] Taiwanese prosecutors accused a senior Nvidia manager, surnamed Chang, of being the “core figure” responsible for the smuggling.[11] Prosecutors alleged that Chang assured managers that on-site inspections had cleared to facilitate approval for the server sale.[12] Prosecutors alleged that the defendants falsified information about the servers’ end users, other uses, and other information to resell them to PRC buyers, sometimes using triangulation shipment methods through Indonesia, Japan, and Hong Kong.[13] Taiwanese customs managed to intercept 56 of the servers, while the other 74 successfully reached PRC customers.[14] The statutory maximum for server-export offenses under Taiwanese law is five years, and prosecutors requested this sentence for four individuals, including senior Nvidia manager Chang.[15]

The indictment follows multiple rounds of investigation. The US Department of Justice unsealed an indictment on March 19 that charged three Super Micro employees for diverting advanced servers to the PRC, including a co-founder and a Taiwan office manager.[16] Taiwanese authorities searched the Taiwanese offices of US firm Super Micro, Super Micro’s Taiwanese distributor Albatron, and data center operator Chief Telecom on June 29 and were investigating six individuals at that time.[17]

US export controls aim to cut off PRC access to advanced chips due to national security concerns. The Nvidia B300 is Nvidia’s most powerful and advanced graphics processing unit (GPU) designed for AI training and inference and is significantly more powerful than the H200 chip that the US has allowed PRC firms to purchase.[18] These export controls have constrained PRC AI development. PLA official newspaper PLA Daily frequently argues that AI is an essential technology for future warfare, with applications ranging from drone swarm capabilities to precision long-range strikes.[19]

Recent PLA discussions of contemporary and future warfare illustrate select lessons learned from the war in Ukraine, including the rapid innovation cycle and importance of legacy systems in modern conflict. The PLA likely continues to view the war in Ukraine as a transitory phase of modern warfare, however, which will give way to “intelligentized” warfare dominated by AI and drone swarm technology. PLA official newspaper PLA Daily published a series of commentaries that demonstrated how select lessons learned from the war in Ukraine are being discussed within the PLA, with a particular emphasis placed on the speed of the innovation cycle in modern combat that will force the PLA to alter its procurement structure to react faster to battlefield changes.[20] A PLA Daily article from August 19 argued that the war in Ukraine demonstrates how battlefield tactics are driving technological development in a new era of warfare.[21] The article focused on how battlefield data rapidly informed the iteration of unmanned systems technology as tactical dynamics changed rapidly on the frontline.[22] An August 16 PLA Daily article similarly discussed the importance of the widespread use of low-cost unmanned systems on the Ukrainian battlefield, describing such systems as the “spear” of modern conflict.[23] Another PLA Daily commentary urged commanders to learn how to integrate legacy platforms into new methods of war, stating that the war in Ukraine has shown that decades-old systems can still be effective on modern battlefields.[24]

Other recent PLA Daily commentaries reflected the PLA’s belief that modern conflict will gradually shift toward a new kind of warfare that centers information technology over physical equipment. An August 19 PLA Daily article argued, for instance, that future AI-enabled drone swarms will revolutionize traditional firepower strike complexes, allowing for total battlefield transparency and extremely rapid precision kill chains.[25] An August 12 PLA Daily article similarly argued that AI will revolutionize long-range reconnaissance, long-range strike, and drone swarm capabilities, reshaping basic battlefield dynamics as they currently exist.[26] An August 13 PLA Daily article clarified these changes, claiming that the modern “war of attrition” would eventually give way to “dissipative warfare.”[27] In dissipative warfare, the center of gravity is an enemy’s operational system rather than their physical equipment and warfighting formations.[28] Dissipative warfare aims to use information superiority, which AI superiority helps create, to prevent an enemy from fighting as a coherent whole.[29]

The PLA’s discussion of the war in Ukraine and the future of warfare suggests that it is preparing for a conflict of a significantly different character while extracting certain tactical lessons from the war in Ukraine. The PLA likely does not view the Ukraine War as an instantiation of dissipative war, as neither Russia nor Ukraine has demonstrated significant AI-enabled swarming technology and algorithm-centric battlefield dynamics, although the role of AI in the war is expanding significantly.[30] This is consistent with ISW-CDOT’s assessment that the PLA likely views the war in Ukraine as a temporary, imperfect phase of future war — or dissipative war — that will only yield certain useful lessons for the war the PLA intends to fight.[31] The lessons that the PLA does extract from contemporary wars, such as the pace of innovation displayed in Ukraine, will likely be shaped by the PLA’s overarching conclusions on the direction of future warfare and the role of emerging technologies. These discussions could suggest that elements of the PLA believe its long-standing conclusions on the informatization and intelligentization of war have largely been confirmed, which could in turn limit the amount of structural and doctrinal change that PLA observations from Ukraine yield.