[Special Commentary on the US-Iran War] ③ The Iran War and the AI Battlefield Revolution: The ‘Paradox of Speed’ and Korea’s Challenges

※ This report is based on publicly available information and reflects the author's personal views.

1. The Great Upheaval of 2026: The Shock of the Iran War

Four years after the shocking Russian invasion of Ukraine, which prompted declarations that the “post-Cold War era is definitively over” (Whitehouse 2022), the international order has once again reached a critical juncture. The operation to capture Venezuelan President Nicolás Maduro in the Caribbean on January 3, 2026, and the “Epic Fury” and “Roaring Lion” operations launched by the United States and Israel against Iran on February 28 are expected to determine not only the future course of warfare but also the direction of the global order.

If the 2022 Russia-Ukraine War was the most serious attack from outside the United States on the liberal order that the US has led since the end of the Cold War, the 2026 Iran War marked the dawn of a new era because the United States, the architect of that order, conducted the war in a manner that conflicted with the rules-based order through actions such as military operations without congressional approval, ambiguous stances on civilian casualties, and the tacit abandonment of international legal standards. While the assessment of historians 100 years from now may differ, the shock inflicted by the Iran War on the existing international political order is judged to be more fundamental than that of 2022.

The Iran War of 2026 requires particularly close analysis because it is the first instance where the United States, the most powerful actor in the international system, compressed the entire decision-making process, from tactics to strategy, through artificial intelligence (AI) decision support. While AI technologies are known to have been utilized in various ways in the Russo-Ukrainian War of 2022 and the Israel-Hamas War of 2023, the level of military AI implemented by the United States in the Iran operations, as discussed below, can be evaluated as a technological and doctrinal innovation that fundamentally alters the paradigm of modern warfare, much like the English longbowmen at the Battle of Agincourt in 1415 or the British army’s ‘Predicted Fire’ at Cambrai in 1917 changed the rules of warfare at the time.

Although a Memorandum of Understanding (MOU) for a ceasefire was signed on June 14, about 100 days after the start of hostilities, a final agreement on core issues has yet to be reached. Nevertheless, understanding the sweeping changes revealed by this war and applying them to the Korean Peninsula context is crucial. The current situation on the Korean Peninsula is grave. Since North Korea’s official declaration of a “hostile, inter-state nation” policy in late 2023, the North Korean nuclear threat has escalated. Following North Korea's dispatch of troops to Russia and the consolidation of military ties between North Korea and Russia, President Xi Jinping visited Pyongyang on June 8 for the first time in seven years, strengthening the North Korea-China strategic relationship and North Korea-China-Russia solidarity. At this juncture, South Korea faces the unprecedented security challenges of both the weakening of the US-led global alliance system and the transfer of wartime operational control.

This issue brief will first examine the fundamental changes brought about by the military use of AI technology in military decision-making systems, as revealed through the course of the Iran War, from the perspective of the ‘paradox of speed.’ To this end, it will assess the extent to which the United States achieved its objectives in the Iran War at tactical and strategic levels and analyze the reasons. Subsequently, it will briefly examine the future direction of US strategy in the Indo-Pacific region, inferred from the US-China summit held from May 13-15, where Trump and Xi Jinping met for the first time since the release of the US National Defense Strategy (NDS) on January 23. Finally, considering these technological and geopolitical changes, this report will propose directions for South Korea’s defense policy.

2. Assessment of the Iran War: The Revolution of Speed and Its Paradox

(1) US Objectives in the Iran War and Current Assessment

Richard Fontaine describes the characteristics of this Iran War as an “anti-Powell Doctrine,” a complete reversal of the US principles for the use of military force established after the Gulf War. This can be summarized as (1) emphasizing ‘flexibility of objectives’ instead of clear goals and a distinct exit strategy; (2) pursuing advantage through ‘Ambiguity and Surprise’ without exhausting non-violent means and issuing ultimatums; (3) omitting congressional approval for the use of force and confirmation of firm public support; (4) pursuing “good enough” outcomes through “Short, sharp military actions” instead of deploying large ground forces; and (5) disregarding the responsibility for post-conflict cleanup (“Pottery Barn rule”) (Fontaine 2026).

The evaluation of the success or failure of military operations is most importantly based on ‘objectives.’ Trump’s constantly shifting targets were considered a unique phenomenon. Nevertheless, to evaluate the outcome of this war, it is necessary to establish, at least at a general level, what the official US objectives were. This paper uses the four core military objectives officially presented by Secretary of War Hagges on March 10 as a benchmark (U.S. Department of War 2026c). First, the destruction of Iran’s ballistic missile and drone capabilities (destruction of launch sites, command and control (C2) nodes, and stockpiles threatening the US and its allies); second, the annihilation of Iran’s naval power (complete elimination of its ability to project power in areas like the Strait of Hormuz and threaten merchant shipping); third, the destruction of Iran’s defense industrial base (DIB) to prevent it from producing and rebuilding missiles, drones, and weapons for several years; and fourth, the permanent prevention of Iran’s nuclear armament.

So, which of these objectives has the United States achieved to date, and what is the likelihood of achieving them in the future? To assess this, it is necessary to analyze separately at the tactical (execution of military operations) and strategic (purpose of military operations) levels.

(2) Tactical Success: The Revolution of Speed

The most notable aspect of this war was the speed of US military operations. A CENTCOM briefing on March 5 stated that over 72 hours, approximately 200 targets in deep Iranian territory, including around Tehran, were struck, resulting in a 90% reduction in Iranian ballistic missile attacks and an 83% reduction in drone attacks after the first day (U.S. Department of War 2026b). By March 10, cumulative strikes reached over 5,000 targets (U.S. Department of War 2026c), and on March 13, Secretary of Defense Hagges announced that US and Israeli air forces combined had struck over 15,000 targets, with airstrikes proceeding at a rate of ‘over 1,000 per day’ (U.S. Department of War 2026d). On April 8, at the time of the ceasefire, Hagges declared that in less than 40 days, “we have dismantled one of the world’s largest armies with less than 10 percent of total US combat power.”

To judge the success of US military operations at the tactical level, General Dan Caine’s report on operations prior to the April 8 ceasefire is crucial. In 38 days of major combat, US forces alone struck over 13,000 targets, of which over 4,000 were “dynamic targets” reflecting real-time intelligence. CENTCOM assessed that approximately 80% of Iran’s air defense systems were destroyed, over 1,500 air defense targets, 450 ballistic missile storage facilities, and 800 disposable attack drone storage facilities were struck, over 2,000 command and control nodes were destroyed, and over 90% of Iran’s regular naval fleet was sunk. It was announced that approximately 90% of Iran’s weapons factories were attacked, over 80% of its missile facilities were eliminated, and nearly 80% of Iran’s nuclear industrial base was hit. Most key leaders, including Supreme Leader Ali Hosseini Khamenei, were eliminated (U.S. Department of War 2026e).

Judging solely by the numbers, this can be seen as an overwhelming tactical victory. The unprecedented operational speed, with an average of one target struck every four minutes over 38 days, appears to have been enabled by the extensive use of AI-based decision support tools. Palantir’s Maven Smart System, developed in collaboration with the Pentagon, appears to have integrated satellite imagery, drone footage, radar data, signals intelligence, cyber intelligence, and open-source information into a unified platform, classifying targets, recommending appropriate weapon systems, and generating strike packages in near real-time. Anthropic’s Claude large language model (LLM) was also embedded in the system, summarizing information, analyzing data, and simulating scenarios, thereby reducing tasks that previously took hours or days to minutes and dramatically compressing the Observe-Orient-Decide-Act (OODA) loop (Lee 2026; Klare 2026). This can be seen as the realization on the actual battlefield of the US Department of War’s declaration in January 2026 of its “AI Acceleration Strategy,” which aimed to remove existing bureaucratic obstacles and integrate cutting-edge frontier AI capabilities into all mission areas to usher in an “era of unprecedented US military AI dominance” (U.S. Department of War 2026a).

Recent experimental research also vividly illustrates this revolution of speed. A study by Dimitrios Doumanas’s team compared the operational planning capabilities of human officers and state-of-the-art LLMs using historical battle scenarios with identifying information removed. The AI models completed analyses that took human officers 1-3 hours in minutes, with the fastest, Claude 3.7, completing the task in an average of 56 seconds, showing a speed over 99% faster than humans. However, human officers sometimes performed better in detailed analyses involving realistic battlefield dynamics such as terrain constraints. Nevertheless, AI outperformed human officers in analyzing the multidimensional ripple effects of tactical decisions and coordinating inter-branch conflicts (Doumanas, Soularidis, and Kotis 2026).

(3) Strategic Failure: Failure to Achieve Political Objectives of Warfare

Despite these impressive figures, most analysts find it difficult to conclude that the Iran War resulted in an “overwhelming US victory” or a mere “case study of military AI innovation.” This is evident when considering the operational objectives presented by the US Department of War: the destruction of Iran’s ballistic missile and drone capabilities, the annihilation of Iran’s naval power, the destruction of Iran’s defense industrial base, and the permanent prevention of Iran’s nuclear armament (Azad 2026; Khan 2026; Cancian and Park 2026).

First, regarding missile and drone capabilities, even the US’s most powerful bunker-buster (GBU-57 MOP) failed to destroy Iran’s underground missile bases, which were built 400-1,500 feet below granite bedrock. Although the US intelligence community claimed to have destroyed 77% of the tunnel entrances, Iran actually preserved 70% of its ballistic missile arsenal before the war and maintained the capability to deploy thousands of missiles by bringing launch platforms out using excavation equipment.

Second, while Iran’s navy was annihilated in the traditional sense, Iran succeeded in controlling the Strait of Hormuz using its “mosquito fleet” of small, fast boats and unmanned surface vessels operated by the Islamic Revolutionary Guard Corps (IRGC), along with its land-based missile systems. Some ships paid passage fees in Chinese yuan to transit the Strait of Hormuz, and countries like India and Pakistan bypassed the US to negotiate directly with Iran, demonstrating the collapse of US control. Consequently, while the direct cost of the war for the US was around $25 billion, the estimated loss to the global GDP due to Iran’s blockade of the Strait of Hormuz ranged from $590 billion to $3.5 trillion.

Third, while some progress can be assessed in destroying Iran’s defense industrial base, positive evaluation of nuclear armament is difficult. Multiple reports suggest that US strikes only delayed Iran’s nuclear program by about six months, failing to completely eliminate it, largely due to the difficulty of directly striking underground facilities. More seriously, the death of Khamenei eliminated the religious edict (Fatwa) prohibiting the development of nuclear weapons, and the collapse of key leadership was filled by hardline IRGC figures, paradoxically strengthening the political justification for pursuing nuclear armament and increasing the likelihood of pursuing much more aggressive policies in the future. Although the MOU did “reaffirm not to acquire or develop nuclear weapons,” as many Republican lawmakers worry, it is highly unlikely that Iran will negotiate the termination of its enrichment program or the dismantling of its nuclear capabilities in good faith.

Fourth, the issue of military costs incurred to conduct the operation, without properly achieving the aforementioned four objectives. Directly, there is the problem of depleting munitions stockpiles. To shoot down an Iranian Shahed drone costing $20,000-$50,000, the US had to endure an extreme cost asymmetry of ‘114 to 1’ by expending a Patriot missile costing $4 million. Furthermore, in terms of overall quantity, although Hagges emphasized that only 10 percent of US forces were used, actual missile inventories are estimated to have been depleted by over 30% overall. Specifically, almost all new ground-launched small precision-guided munitions (PrSM) were expended, and over half of THAAD and Patriot systems, for which there are no substitutes, were depleted (Cancian and Park 2026). Due to the limitations of the US defense industry’s production capacity, rebuilding these depleted missiles is expected to take at least one to four years, consequently creating a significant structural vulnerability in the US’s core strategy of deterring China in the Indo-Pacific region.

Fifth, beyond military costs, there are political costs. The war’s initiation without prior consultation with European allies caused a severe diplomatic rift with NATO, leading to non-cooperation from European allies such as Spain, France, Germany, and the UK regarding airspace transit and base usage during the actual conduct of the war. Conversely, the anti-US coalition extending from China, Russia, North Korea, to Iran was consolidated. The easing of sanctions on Russian oil and the increased effect of rare earth export controls during the rebuilding of weapons systems significantly strengthened the strategic positions of Russia and China. Domestically, public opinion turned sharply against the administration due to soaring oil prices and the prolonged war, causing approval ratings to plummet into the 30% range.

(4) The Paradox of Speed

Tahir Azad points out that with this war as a turning point, the era of “Strategic Abundance” that the US has enjoyed since 1991 has ended, and the US has now fallen into a state of “Strategic Insolvency.” Ultimately, the lesson of the Vietnam War—that tactical success does not automatically guarantee the achievement of strategic and political objectives—is being repeated in the age of AI. Some assess that this war “will be far more prolonged and severe than the failure in Vietnam” (Musgrave 2026). So why did the revolution of speed, far exceeding human cognitive abilities, and the resulting immense tactical success lead to strategically disastrous outcomes?

Edward Luttwak explains that the realm of strategy is governed not by everyday linear logic but by a “paradoxical logic” where opposing elements combine and invert. A prime example is the need to abandon flat terrain for difficult routes that the enemy would never expect, a logistically disastrous choice, in order to achieve a surprise attack. Success, once it passes the “culminating point of victory,” is bound to reverse, and “a tactically good path can be a strategically bad path.” The German army’s ill-advised advance toward Moscow during World War II is a prime example.

Luttwak’s ‘paradoxical logic’ has significant implications for the era of military AI revolution. This is the ‘paradox of speed’ emphasized in this issue brief. First, excessive speed weakens human control. In a rapidly unfolding battlefield, when faced with AI recommendations derived from analyzing vast amounts of multi-dimensional data beyond human imagination, it becomes difficult to thoroughly consider the accuracy and military validity of identified targets, the possibility of civilian casualties, and strategic proportionality. Furthermore, it is very difficult to be confident that one’s own judgment is superior to the AI’s. In this context, if the entire process from detection to analysis to strike is accelerated by AI-enabled ‘compression of decision-making,’ the human supervisor’s oversight function, even if present procedurally, can become practically meaningless (Csernatoni 2026). The Israeli Lavender system is even reported to have used an average of only 20 seconds for human analysts to review each AI-recommended target. What level of prudence can be exercised in 20 seconds? Although CENTCOM emphasized that human supervisors were always “human-in-the-loop” during the Iran War, the fact that thousands of dynamic targets were processed daily suggests that maintaining effective human control amidst the revolution of speed is an extremely difficult task.

Second, there are structural limitations of AI models. AI models are not free from biases and errors in their training data, and due to the nature of machine learning, they can fall into serious errors when making judgments in situations with insufficient data. Nevertheless, the decision-making process of AI often remains opaque, like a black box. The tragic incident of a girls' primary school in Minab, southern Iran, being hit on February 28, drew media attention as a case suggesting the possibility of errors in AI targeting systems. The excessive use of AI-based targeting can lead to unintended fatal errors when machine-based statistical models are executed without meaningful human control (Amaral 2026). War, in particular, is a complex phenomenon involving political objectives, civilian casualties, and asymmetric enemy responses, making it difficult to objectively assess the success of AI-based military operations prior to their execution based solely on speed or accuracy metrics.

Third, there are serious errors arising from the interaction of the two factors mentioned above. The human brain, regardless of the quality or truthfulness of information, tends to assign “fallacy of overconfidence” to more information. When a machine, after analyzing vast amounts of data collected from various platforms, proposes courses of action through various statistics, probabilities, and visualized displays, humans are prone to the fallacy of overconfidence, easily considering it more reliable than their own limited judgment. When this encounters a time-constrained decision-making environment, such as during a crisis or war, the consequences can be catastrophic. This creates a feedback loop where the first problem exacerbates the second, and the second problem reinforces the first.

This paradox of speed is confirmed in writing in the MOU signed on June 14. Despite striking over 13,000 targets and eliminating Iran’s leadership, the conditions the US accepted were the phased lifting of all sanctions, reconstruction aid totaling at least $300 billion, the lifting of the blockade and withdrawal of troops, and the de facto continuation of nuclear enrichment capabilities. Tactical climax led to strategic concession (Foreign Policy 2026).

3. US Strategic Realignment and Korea’s Challenges

The MOU examined above holds another significance for the United States. Its strategic essence lies in the US’s attempt to return to a ‘priority-based strategy.’ The Trump administration will likely seek to reduce military and political burdens in the Middle East and reallocate resources and attention to the Indo-Pacific and containment of China. This is a natural consequence in line with the direction set forth in the National Defense Strategy (NDS) released on January 23: defense of the homeland and Western Hemisphere, containment of China through denial and deterrence in the first island chain, expanded burden-sharing with allies, and strengthening the defense industrial base (Yang-gyu Kim 2026a; 2026b).

However, the paradox of speed revealed by the Iran War casts a shadow over this realignment plan. The recovery of munitions, missile defense, and aircraft carrier capabilities depleted by the Iran War will take time. Therefore, the post-ceasefire phase is likely to be accompanied by a period of ‘window of vulnerability’ where there is a gap between the US’s intention to refocus on the Indo-Pacific and its actual available military power, rather than immediate power recovery. The US-China summit in May illustrates this structure well. President Trump met with President Xi Jinping burdened by the Iran War, oil price pressures, the need for diplomatic achievements before the midterm elections, and depleted munitions. China, on the other hand, entered the negotiations with its key leverage—rare earths, critical minerals, supply chains, and enhanced military capabilities around the Taiwan Strait—intact. As a result, the US achieved some tangible outcomes in areas such as energy and agricultural purchases, but issues concerning Taiwan, semiconductors, and North Korea were effectively pushed to the periphery of the agenda. The fact that the US could not provide adequate responses to China’s strong demands indirectly indicates which side holds the advantage in US-China negotiations. The display of overwhelming military power through the Iran War did not directly translate into increased leverage in negotiations with China.

In this context, the discussion in the ROK-US alliance is likely to shift from ‘burden-sharing’ to ‘capability-sharing.’ The US may demand broader roles from South Korea, beyond mere cost-sharing, including shipbuilding, maintenance, repair, and overhaul (MRO) of naval vessels, munitions production, missile defense, intelligence sharing, rear bases, and even support in the event of a Taiwan contingency. The risk is that South Korea may become a peripheral theater supporting US strategy against China, rather than focusing on the defense of the Korean Peninsula. While South Korea must acknowledge the need to modernize the ROK-US alliance, it must clearly define its objective as ‘strengthening strategic stability on the Korean Peninsula,’ not ‘supporting US deterrence in the Indo-Pacific,’ and demand a balance between the military risks it must bear and the corresponding strategic rewards.

The post-ceasefire phase also has dual implications for the North Korean issue. President Trump may seek to leverage the North Korean issue again to highlight himself as a “peacemaker” and showcase diplomatic achievements before the midterm elections, following the Iran War. However, North Korea in 2026 is different from North Korea in 2018. North Korea has advanced its nuclear capabilities, constitutionally enshrined its status as a nuclear-weapon state, explicitly rejected denuclearization, redefined inter-Korean relations as “hostile, inter-state,” and is increasingly adopting an attitude of excluding South Korea from negotiations. In particular, North Korea, having observed that even Iran, with its significant nuclear potential and missile capabilities, could not withstand a large-scale US military attack, is likely to focus on enhancing the survivability and dispersal of its nuclear forces and improving its tactical nuclear operational capabilities, rather than denuclearization. In this situation, any US-North Korea discussion of “stable coexistence” (Aum and Panda 2025) conducted without South Korea’s involvement could devolve into a “dangerous bargain” from Seoul’s perspective. South Korea does not need to oppose North Korea-US dialogue or risk mitigation measures themselves, but it must prevent the US from using aspects such as de facto recognition of North Korea’s status as a nuclear-weapon state, ROK-US combined exercises, strategic asset deployments, sanctions relief, or a declaration of the end of war as bargaining chips with North Korea without prior consultation with South Korea.

Ultimately, the most important lesson the Iran War offers for the Korean Peninsula is that superiority in speed does not inherently guarantee strategic success. South Korea has already embarked on integrating AI into its targeting, surveillance and reconnaissance, decision support, and kill chain systems to counter the escalating North Korean nuclear threat, and this direction is unavoidable. However, on the Korean Peninsula, with its short depth and extremely compressed decision times, North Korea may perceive the strengthening of the South Korean military’s AI-based kill chain capabilities as a preemptive neutralization attempt against its nuclear assets. Amidst mutual distrust, both sides may feel the pressure of a “use-it-or-lose-it” scenario, where striking first is advantageous (Kim 2025). Therefore, the South Korean military’s AI should contribute not to ‘first-strike capability’ but to ‘the ability to survive even after being hit, rapidly restore command and control, and respond with certainty.’ Strategic stability on the Korean Peninsula in the AI era can only be achieved when the resilience of command and control and the certainty of retaliation are enhanced, and when a crisis management mechanism is institutionalized for both South Korea and the US to jointly assess and manage the signals sent to North Korea by AI-based targeting and decision support, and the pathways to unintended escalation (Kim 2025; Yang-gyu Kim 2026c). In an era where speed does not guarantee victory, what we must pursue is not superiority in the speed race, but a robust deterrence system that incorporates prudence. ■

References

Amaral, Nilza. “The Iran War Highlights the Creeping Use of AI in Warfare.” Chatham House (March 27, 2026).

Aum, Frank, and Ankit Panda. “Pursuing Stable Coexistence: A Reorientation of U.S. Policy Toward North Korea.” Carnegie Endowment for International Peace, April 2025.

Azad, Tahir. “The Cost of the Iran War for the United States: A Strategic Blunder in Five Dimensions.” Small Wars Journal, April 27, 2026.

Cancian, Mark F., and Chris H. Park. “Last Rounds? Status of Key Munitions at the Iran War Ceasefire.” Center for Strategic and International Studies (CSIS), April 21, 2026.

Csernatoni, Raluca. “The Fog of AI War.” Carnegie Europe (April 16, 2026).

Doumanas, Dimitrios, Andreas Soularidis, and Konstantinos Kotis. “Causal Reasoning and Large Language Models for Military Decision-Making.” AI Vol. 7, No. 1 (2026), Article 14.

Fontaine, Richard. “Trump’s Way of War.” Foreign Affairs, March 2, 2026.

Khan, Bilal. “US-Iran War 2026: Operation Epic Fury and the Hormuz Crisis.” Quwa, May 3, 2026.

Kim, Yang Gyu. “Building Conditions for Peace on the Korean Peninsula: Prioritizing Countervalue over Counterforce.” International Journal of Korean Unification Studies 34(2): 1–28, 2025.

Klare, Michael. “AI Plays Major Role in the War on Iran.” Arms Control Today, May 1, 2026.

Lee, Daniel. “AI Central in US-Iran 2026 War as Targeting Systems Speed Strikes but Spark Accuracy and Ethics.” International Business Times, April 5, 2026.

Musgrave, Paul. “Iran Is a Bigger Defeat Than Vietnam.” Foreign Policy, June 16, 2026, 3:04 PM

U.S. Department of War. “War Department Launches AI Acceleration Strategy to Secure American Military AI Dominance.” Immediate Release, January 12, 2026a.

U.S. Department of War. “Secretary of War Pete Hegseth and Admiral Brad Cooper, Commander of U.S. Central Command, Hold a Press Briefing at Central Command Headquarters on U.S. Military Operations in the Middle East.” Press briefing transcript, March 5, 2026b.

U.S. Department of War. “Secretary of War Pete Hegseth and Chairman of the Joint Chiefs of Staff Gen. Dan Caine Hold a Press Briefing.” Press briefing transcript, March 10, 2026c.

U.S. Department of War. “Secretary of War Pete Hegseth and Chairman of the Joint Chiefs Air Force Gen. Dan Caine Hold a Press Briefing.” Press briefing transcript, March 13, 2026d.

U.S. Department of War. “Secretary of War Pete Hegseth and Chairman of the Joint Chiefs Air Force Gen. Dan Caine Hold a Press Briefing.” Press briefing transcript, April 8, 2026e.

U.S. Department of War. 2026 National Defense Strategy. January 23, 2026.

Foreign Policy, “U.S.-Iran Memorandum of Understanding: Full Text.” June 17, 2026.

Xinhua. “Chinese, U.S. Presidents Agree on New Vision for Bilateral Ties in Beijing Talks.” May 14, 2026.

Kim, Yang Gyu. “Changes in Sino-US Military Security Strategy in 2026 and South Korea's Defense Strategy.” *EAI New Year Planning Special Commentary Series* 8, January 22, 2026.

Kim, Yang Gyu. “The Paradox of the 'Primary Responsibility' Claim and 'Limited Support': The 2026 NDS and South Korea's Choice.” *RINSA Security Review* No. 236, Institute for National Security Strategy, February 28, 2026.

Kim, Yang Gyu. 2026c. “AI Battlefield Revealed by the Middle East War: Challenges for South Korean Defense.” *Kukbang Ilbo*, May 19, 2026.

■ Kim Yang-gyu_Professor, Korea National Defense University

■ Responsible for and Edited by: Lee Sang-jun_EAI Research Fellow
    Inquiries: 02 2277 1683 (ext. 211) | leesj@eai.or.kr