What happened between Anthropic and the U.S. Department of Defense?

In July 2025 the two signed a contract worth up to $200 million, but they ruptured in 2026 over issues including fully autonomous weapons. The Pentagon designated Anthropic a supply chain risk; Anthropic sued in March 2026, and that same month a California federal court issued a preliminary injunction blocking the designation. On May 1, 2026 the Pentagon signed contracts for classified networks with eight companies excluding Anthropic.

What is the current status of the U.S. AI Diffusion Rule?

Promulgated in January 2025 and scheduled to take effect on May 15, the rule was rescinded on May 13 — two days before it was to take effect. As a replacement, the U.S. Commerce Department issued guidance that using certain Chinese advanced computing ICs such as Huawei's Ascend could be subject to General Prohibition 10 (GP10), along with warnings against circumvention (diversion).

How many nuclear warheads currently exist worldwide?

The Federation of American Scientists (FAS) estimates about 12,187 warheads worldwide as of early 2026. Russia (~5,420) and the United States (~5,042) hold the vast majority, while China has about 620 and is analyzed to be adding roughly 100 warheads per year.

Why are AI and semiconductors subject to export control?

Because AI models and advanced semiconductors are dual-use technologies usable for both civilian and military purposes. The same chips and models can power generative AI services and be diverted to autonomous weapon control, cyberattacks, and the simulation of nuclear and missile development — which is why countries incorporate them into export controls.

AI and the Military — The New Frontier of National Security

Introduction — Purpose of This White Paper

This is Hamamoto from TIMEWELL.

TIMEWELL provides services that underpin export control across global supply chains. The risk that advancing cutting-edge technology is inadvertently diverted to military use is no longer something either international trade or national security can afford to ignore. AI and advanced semiconductors in particular are the archetype of "dual-use" (civilian-and-military) technology — the same chips and models can power everyday consumer services like generative AI, and can just as readily run the autonomous control of a weapon or the simulation of nuclear and missile development. The tug-of-war over export control in recent years increasingly converges on this single point.

This white paper organizes how AI and other advanced technologies are used in the military domain and what impact they have on the global balance of power and our daily lives, then connects the analysis in a single line to export control — the front line of technology management. The first edition was published in February 2026, but we have fully rewritten it in July 2026 to reflect major developments since then: the rupture between Anthropic and the U.S. Department of Defense (DoD), the latest estimates of nuclear forces, and the rescission of the U.S. AI Diffusion Rule. If you want a quick check on whether the products or technologies your company handles carry a risk of military diversion, you can start with our export control classification check.

Chapter 1: Introduction — The Paradigm Shift in Warfare

1-1. Redefining AI and the Military

When discussing "AI and the military" today, many people might picture drones that fly autonomously and strike targets, or the robot soldiers of science fiction films. But the essence goes well beyond simply automating and autonomizing weapons. The transformation that artificial intelligence (AI) brings to the military is a paradigm shift in intelligence itself — one that shakes every domain of security to its foundations: national decision-making, intelligence analysis, strategic planning, logistics, cyber defense, and even public perception.

AI can instantaneously analyze volumes of data no human could ever process, forecast the future course of a battle, and present the options with the highest probability of success. It can complete in minutes the operational planning that once took a general staff weeks, catch subtle indicators that skilled pilots would miss, and neutralize enemy cyberattacks in real time. AI has arrived not as an "automation" tool that replaces the limbs of soldiers, but as a "new intelligence" that extends the commander's mind and transforms national decision-making itself. It is no exaggeration to say that whoever masters this intelligence will master interstate competition in the 21st century.

1-2. Structure of This White Paper

This white paper begins by reviewing Japan's military strategic history from the Kamakura samurai to the prewar intelligence state (Chapter 2), then surveys the wisdom of classical military strategy from Sun Tzu to Clausewitz (Chapter 3). It goes on to cover the history and cutting edge of nuclear weapons (Chapter 4), the generational evolution of fighter aircraft (Chapter 5), and the key military technologies of land, sea, and air (Chapter 6), before examining how the balance of power among the U.S., China, and Russia has crumbled and how China is strengthening its military with AI (Chapter 7). It then explores how modern warfare has transformed into "hybrid warfare" (Chapter 8), the tension between AI companies and the state (Chapter 9), the threats of cyber and cognitive warfare (Chapter 10), directed energy weapons and covert operations (Chapter 11), fully autonomous weapons and robot soldiers (Chapter 12), AI military simulation (Chapter 13), the militarization of space (Chapter 14), and the contest over energy and power dominance (Chapter 15). On that basis, it organizes the contest over "dual-use technology and export control" that underlies all of the above, using the latest regulatory picture (Chapter 16), and closes with strategic recommendations for protecting Japan's national interests (Chapter 17).

Chapter 2: Japan's Military Strategic History — From Kamakura Samurai to the Intelligence State

2-1. Kamakura Samurai and the "Way of Bow and Horse"

Tracing the roots of Japan's military strategy leads to the "samurai" — the warrior class established from the late Heian period through the Kamakura period. Kamakura samurai in particular were known for their austere, robust spirit and high combat skills, with mounted archery — the "way of bow and horse" — at the center of their fighting art. The ability to shoot accurately from horseback was the fruit of years of training and symbolized the high combat capability of the individual warrior. During the Mongol invasions of the late 13th century, Kamakura samurai clung to the traditional style of single combat and struggled against Mongol forces that used massed formations and gunpowder weapons known as "tetsuhau." This experience — Japan's first large-scale foreign war — became the occasion for recognizing the importance not only of individual valor but of organized combat.

2-2. The Information Warfare and Ninja of the Warring States Period

The roughly 150 years of the Warring States period following the Onin War was the era in which information warfare was most developed in Japanese history. Each daimyo maintained specialist groups to gather intelligence on rival domains and devise stratagems — the most representative being the "ninja." Based in Iga (Mie Prefecture) and Koka (Shiga Prefecture), these operatives were skilled in disguise, infiltration, fire techniques, and psychological warfare. They slipped into enemy castles to steal information, and at times carried out assassinations and sabotage. Takeda Shingen's "three-person agents," the Hojo clan's "Fuma party," and others waged information warfare using their own intelligence organizations. In this era, the understanding spread that "eight-tenths of victory or defeat is decided before the battle begins" — a deep appreciation of just how vital information is.

2-3. Prewar Geopolitical Research and Global Intelligence Operations

After the Meiji Restoration, as Japan pursued "a rich country and a strong army" to stand alongside the Western powers, it also awoke to the importance of modern intelligence operations. The Russo-Japanese War (1904–1905) was a landmark event in which Japanese intelligence activity had a major impact on world history. Colonel Motojiro Akashi, then military attaché at the Japanese legation in Russia, used a vast operational fund equivalent to roughly 1% of the national budget to support anti-government forces and revolutionaries inside Russia. He succeeded in a grand scheme to sow chaos within Russia and sap its ability to continue the war from within. This "Akashi Operation" is credited with contributing greatly to Japan's victory.

Prewar Japan also studied "geopolitics" — a discipline born in Germany — as an important field for thinking about national strategy. Scholars such as Sadashige Komaki at Kyoto Imperial University advocated a "Japanese geopolitics" based on Japan's geographical characteristics and historical mission, which became one of the theoretical pillars of the "Greater East Asia Co-Prosperity Sphere" concept. At spy-training institutions typified by the Imperial Army's Nakano School, geopolitics was taught as a required subject alongside espionage and sabotage techniques, and graduates scattered around the world to engage in intelligence collection and covert operations. In this way, prewar Japan possessed a world-class intelligence network and geopolitical thinking.

Chapter 3: The Wisdom of Classical Military Strategy — From Sun Tzu to Clausewitz

The history of war is the history of humanity itself, and the military strategies and philosophies cultivated within it have determined the rise and fall of nations across the ages. To understand the revolutionary changes AI is bringing to the military sphere, we must first survey the core of classical military thought and understand how it connects to the present.

※ This chapter provides an overview of the world's classical military thought, as a companion to Chapter 2 on Japan's military strategic history.

3-1. Sun Tzu's Art of War — Intelligence and Stratagem for Winning Without Fighting

The Art of War, born in ancient China's Spring and Autumn period roughly 2,500 years ago, is the world's oldest and most influential military treatise. Said to be authored by Sun Wu, its thirteen chapters treat war not as a mere clash of force but as a complex phenomenon in which politics, economics, diplomacy, geography, and psychology are all intertwined.

The core of Sun Tzu's thought is the ideal of non-warfare: "To subdue the enemy without fighting is the acme of skill." Armed conflict is a last resort; the highest achievement is to frustrate the enemy's intentions through diplomacy and stratagem and claim victory without battle. Indispensable to this is "intelligence." As the famous line — "Know the enemy and know yourself, and in a hundred battles you will never be in peril" — shows, Sun Tzu emphasizes accurately grasping the realities of both sides and forming plans based on careful analysis. He also states that "all warfare is deception," explaining the effectiveness of deception, surprise, and diversion in seizing the initiative and confusing the enemy into defeat.

AI technology holds the potential to realize Sun Tzu's ideals at an unprecedented level. AI's high-speed analysis of massive datasets becomes a powerful tool for "knowing the enemy and knowing yourself," enabling the instant discovery of enemy weaknesses and intentions. AI-generated fake news and the automation of cyberattacks can become the means to execute "deception" at greater sophistication and scale. AI is the technology that resurrects Sun Tzu's art of war for the modern age, across 2,500 years.

3-2. Napoleon's Military Revolution — The Birth of the Conscript Army and Total War

Napoleon Bonaparte, who emerged from the turmoil of the French Revolution, swept across all of Europe through his military genius. His strategy overturned the conventions of the limited wars of the monarchical era from the ground up and opened the door to modern warfare.

The source of Napoleon's strength lay in three revolutionary elements. First, the conscript army that the French Revolution produced — this enabled the formation of large national armies burning with patriotism, overwhelming the previously mercenary-based armies in sheer numbers. Second, the creation of the corps system: each corps was a self-contained mini-army with infantry, cavalry, and artillery, enabling large forces to maneuver flexibly and rapidly. Third, exploiting this mobility through interior lines and the concentration of force: when surrounded by multiple enemies, positioning centrally to defeat them one by one, and concentrating force at the decisive point of battle to annihilate the enemy's main body — these were his specialty.

In the AI age, the principles of mobility and concentration of force that Napoleon pursued undergo further evolution. AI analyzes countless variables — battlefield conditions, logistics, enemy dispositions — in real time and computes optimal movement routes and attack timing. AI autonomously operating formations of drones and unmanned weapons, concentrating force at the most vulnerable point of the enemy's defenses at the optimal moment — precisely the tactics Napoleon idealized — are being realized through AI.

3-3. Clausewitz's On War — The Unchanging Relationship Between War and Politics

Drawing on his experience of the Napoleonic Wars, the Prussian soldier Carl von Clausewitz produced in On War an enduring masterwork of profound philosophical insight into the nature of war.

On War's most famous proposition is: "War is nothing but the continuation of politics by other means." This shows that war does not exist for its own sake but is always an instrument for achieving political purposes. Clausewitz also explained the nature of war as a "strange trinity": ① the blind impulse of "hatred and animosity" in the people; ② the "play of chance and probability" in the commander and army; and ③ the "reason" of the government as a political instrument. He further advanced the concept of "friction" — all the unexpected obstacles that arise between plan and execution (bad weather, intelligence errors, troop fatigue, and so on) — arguing that this is the fundamental reason real war never proceeds according to theory.

Even with AI introduced into war, Clausewitz's fundamental principle that "war is an instrument of politics" remains unshaken. However, if the speed of AI decision-making outpaces the cycle of human political judgment, there emerges the risk of a "flash war" in which politics can no longer control war. And while AI may reduce "friction," it may also generate new AI-era frictions: enemy cyberattacks against the AI, algorithmic flaws, and the "black box problem" of AI taking unpredictable actions.

3-4. Mahan's Theory of Sea Power — He Who Commands the Sea Commands the World

The American naval officer Alfred Thayer Mahan, in his The Influence of Sea Power upon History (1890), argued how the rise and fall of nations throughout history has depended on "sea power," laying the foundations of geopolitics.

Mahan argued that the ocean is a vast highway connecting the world, and that commanding it (command of the sea) is indispensable to the protection of commerce and national prosperity. He identified six elements enabling a nation to establish sea power: geographical position, physical conformation, extent of territory, population, national character, and the character of the government. Mahan's theory profoundly influenced the American leadership of his time, including President Theodore Roosevelt, and became the impetus for the United States to steer toward becoming a great naval power.

Even today, more than 90% of world trade depends on maritime transport, and the importance of sea power is unchanged. AI technology transforms the very nature of the competition over sea power. AI-equipped unmanned surface vessels (USV) and unmanned underwater vehicles (UUV) dramatically enhance the ability to monitor vast oceans around the clock and detect and track enemy submarines and ships. One could say a new seventh element — "AI capability" — has been added to the components of sea power Mahan described.

3-5. Lessons for the AI Age from Classical Military Thought

Sun Tzu's information warfare, Napoleon's war of maneuver, Clausewitz's relationship between politics and war, Mahan's command of the sea. Viewed through the new lens of AI, these classical military philosophies reveal to us both their universal value and the aspects that change with the times.

Classical thinker	Core principle	Significance in the AI age
Sun Tzu	Information superiority, non-warfare, deception	Theoretical basis for AI big-data analysis and cyber/cognitive warfare
Napoleon	Mobility, concentration of force, war of annihilation	Ultra-fast maneuver and optimal force deployment by AI autonomous weapon swarms
Clausewitz	War as an instrument of politics; friction	Risk of AI runaway (flash war) and new forms of "friction"
Mahan	Sea power, command of the sea	Maritime dominance by AI-equipped unmanned vessels; the addition of "AI capability"

What these philosophies teach is that however much technology advances, the essence of war — information, maneuver, political purpose, and control of domains — does not change. AI becomes the means to optimize and realize these classical principles to the extreme; but at the same time, it confronts us with the risk of conflict escalating beyond human control, and with profound ethical questions.

Chapter 4: The History and Cutting Edge of Nuclear Weapons — The Apex of Destruction and the Logic of Deterrence

4-1. From the Manhattan Project to Hiroshima and Nagasaki

In 1942, during World War II, the United States, fearing Nazi Germany's atomic bomb development, launched the top-secret "Manhattan Project." Many brilliant scientists were mobilized, centered on Los Alamos National Laboratory, and on July 16, 1945, the world's first nuclear test (the Trinity test) succeeded at Alamogordo, New Mexico. Then, on August 6 of that year, the uranium-type atomic bomb "Little Boy" was dropped on Hiroshima, and on August 9 the plutonium-type atomic bomb "Fat Man" on Nagasaki. Both cities suffered catastrophic destruction and hundreds of thousands of precious lives were lost. These two wartime uses made the inhumanity of nuclear weapons known to the world while, at the same time, signaling that their overwhelming destructive power announced the dawn of a new era.

4-2. The Cold War and Mutually Assured Destruction (MAD)

The atomic bombings of Hiroshima and Nagasaki delivered a tremendous shock to the Soviet Union, and Stalin accelerated his own country's nuclear development. In August 1949 the USSR succeeded in its first nuclear test, and America's nuclear monopoly came to an end. From this point, the U.S. and the USSR entered an era of "nuclear arms racing," competing to develop and deploy ever more powerful and diverse nuclear weapons while stoking mutual distrust and fear.

As the arms race advanced, both the U.S. and the USSR came to possess the ability to absorb a first strike from the other and reliably carry out a retaliatory attack (second-strike capability). This created a situation in which "if one side uses nuclear weapons, the other's retaliation devastates both, and there is no victor." This theory, called Mutually Assured Destruction (MAD), ironically played a certain role in deterring the outbreak of nuclear war. During the 1962 Cuban Missile Crisis the world stood at the brink of nuclear war, but the logic of MAD ultimately guided the leaders of both nations toward restraint.

4-3. The NPT and the Current State of Nuclear Arsenals

An increase in the number of nuclear-armed states would dramatically raise the risk of nuclear war. From this concern, the Treaty on the Non-Proliferation of Nuclear Weapons (NPT) was adopted in 1968 and entered into force in 1970. The NPT recognizes five states — the U.S., Russia (the USSR), the UK, France, and China — as "nuclear-weapon states" and permits their possession, while prohibiting all other non-nuclear-weapon states from acquiring nuclear weapons and imposing on the nuclear-weapon states an obligation to negotiate nuclear disarmament in good faith.

As of early 2026, beyond the five great powers recognized by the NPT, India, Pakistan, and North Korea have declared possession of nuclear weapons, and Israel's possession is considered virtually certain. According to the Federation of American Scientists (FAS), the total number of nuclear warheads worldwide is about 12,187. Although this is far below the Cold War peak, the pace of reduction has slowed, and some countries have begun to increase their arsenals.

Country	Total stockpile (estimated)	Deployed (estimated)
Russia	5,420	1,796
United States	5,042	1,770
China	620	34
France	370	280
United Kingdom	225	120
India	190	0
Pakistan	170	0
Israel	90	0
North Korea	60	0

Source: Federation of American Scientists (FAS), "Status of World Nuclear Forces" (early-2026 estimate; reference point March 2025)

China stands out among these. Having long professed a policy of "minimum deterrence," China now increases its warheads at a pace of roughly 100 per year, and its deployed strategic warheads rose from 24 the previous year to 34. Analyses by FAS and the Bulletin of the Atomic Scientists (July 2026 issue) project that its operational warheads will reach about 1,000 by 2030 and up to 1,500 by 2035. China continues to refuse to join the U.S.-Russia arms control framework, and its opacity heightens the international community's concern.

4-4. New Trends in Nuclear Weapons Technology

Since the end of the Cold War, even as nuclear disarmament has progressed, the modernization of nuclear weapons and the drive to develop new technologies have grown more active.

Hypersonic weapons: Flying at hypersonic speeds of Mach 5 or above along irregular low-altitude trajectories, "hypersonic weapons" are considered extremely difficult to detect and intercept with current missile defense systems. In particular, "hypersonic glide vehicles (HGVs)" — which separate from a ballistic missile and then glide — are the object of intense development competition among the U.S., China, and Russia.

Small tactical nuclear weapons: The development of "small tactical nuclear weapons" with suppressed explosive yields, intended for use in limited conflicts, is also progressing. These lower the threshold for using nuclear weapons and carry the danger of inducing escalation as "usable nuclear weapons." Russia in particular is said to emphasize an "escalate to de-escalate" strategy that would use tactical nuclear weapons if it fell into an inferior position in conventional forces.

Nuclear fusion weapons research: Fusion weapons, exemplified by the hydrogen bomb, possess destructive power incomparable to atomic bombs. Nations use simulation technology and other means to maintain and improve the performance of their nuclear weapons, and it has been pointed out that fusion research technology could be applied to weapons development.

4-5. The Nuclear Development Problems of North Korea and Iran

Outside the NPT regime, the nuclear development of North Korea and Iran has become an urgent challenge for the international community. North Korea withdrew from the NPT and forced through multiple nuclear tests. It has repeatedly conducted ballistic missile launch tests and is advancing the development of ICBMs capable of reaching the U.S. mainland. Iran concluded a nuclear agreement (the JCPOA) with six nations including the U.S. and Europe in 2015, but the U.S. unilaterally withdrew in 2018. Since then, Iran has raised its uranium enrichment levels and other measures, reigniting concerns about nuclear weapons development.

Chapter 5: The Evolution of Fighter Aircraft — From the 1st to the 6th Generation, and On to the AI Pilot

5-1. The Generational Evolution of Jet Fighters

Fighter aircraft, as the star of airpower, have been developed by gathering the finest of the most advanced technology of their era. The evolution of jet fighters is organized by the concept of generations.

1st generation (1950s): Subsonic jets. The Korean War saw the first jet-versus-jet aerial combat. Representative aircraft: F-86 Sabre, MiG-15.

2nd generation (1950s onward): Gained supersonic flight capability and began carrying radar and air-to-air missiles. Combat shifted to beyond visual range (BVR). Representative aircraft: F-4 Phantom II, MiG-21.

3rd generation (1960s onward): Under the doctrine of missile supremacy, emphasized missile armament over dogfighting performance. However, experience in the Vietnam War prompted a reassessment of the importance of close-in dogfighting.

4th generation (1970s onward): Designed with high dogfighting performance as a priority. Characterized by high mobility from powerful engines and advanced control systems such as fly-by-wire. Representative aircraft: F-15 Eagle, F-16 Fighting Falcon, Su-27 Flanker.

5th generation (2000s onward): The defining feature is stealth. By carefully shaping the airframe and using radar-absorbent materials, radar cross-section (RCS) is dramatically reduced, realizing "First Look, First Shot, First Kill." Representative aircraft: F-22 Raptor, F-35 Lightning II, Su-57, J-20.

Generation	Era	Main characteristics	Representative aircraft
1st	1950s	Subsonic jets	F-86, MiG-15
2nd	1950s onward	Supersonic, missile-armed	F-4, MiG-21
3rd	1960s onward	Missile emphasis	F-104, MiG-23
4th	1970s onward	Dogfighting emphasis, FBW	F-15, F-16, Su-27
4.5th	1990s onward	AESA, limited stealth	Rafale, Typhoon
5th	2000s onward	Stealth, sensor fusion	F-22, F-35, J-20

5-2. The Competition to Develop 6th-Generation Fighters

Today, the competition to develop 6th-generation fighters is intensifying across the world's nations. Concepts common to 6th-generation aircraft include stealth surpassing the 5th generation, the use of AI, coordination with unmanned aircraft (manned-unmanned teaming), and adaptation to network-centric warfare.

NGAD (Next Generation Air Dominance): The next-generation air dominance program the U.S. is pursuing. It is characterized by a "System of Systems" concept that links crewed fighters with accompanying unmanned Collaborative Combat Aircraft (CCA).

GCAP (Global Combat Air Programme): A program jointly developed by Japan, the UK, and Italy. It integrates Japan's F-X and the UK's Tempest program and aims to begin deployment in 2035. At the end of 2024 the three countries agreed to establish the intergovernmental organization GIGO to oversee joint development, and in 2025 the building of the organizational structure got into full swing. Saudi Arabia's participation has also been discussed.

FCAS (Future Combat Air System): A future combat air system jointly developed by France, Germany, and Spain.

5-3. The Birth of the AI Pilot — DARPA's ACE Program

The ACE (Air Combat Evolution) program run by the U.S. Defense Advanced Research Projects Agency (DARPA) is researching the automation of aerial combat by AI. In 2020, AI produced results overwhelming a veteran pilot in simulation, and in 2023 the AI-equipped experimental aircraft X-62A succeeded in a simulated dogfight against a human-piloted F-16.

In the future, a style may become mainstream in which the pilot fights as the commander of the battle, leading multiple AI-piloted unmanned aircraft. Freed from the physical limits (G-forces) that human pilots experience, unmanned combat aircraft will be capable of high maneuvers surpassing crewed aircraft, fundamentally changing the nature of aerial combat.

Chapter 6: Key Military Technologies of Land, Sea, and Air — The Platforms Supporting Modern War

6-1. Key Technologies on Land

The evolution of main battle tanks (MBT): Born to break the trench deadlock of World War I, the tank became the star of blitzkrieg in World War II. Today, Generation 3.5 tanks (M1A2 SEP, the Type 10, and others) with enhanced networking and urban-warfare capability are mainstream. The latest technologies under development include active protection systems (APS) that physically intercept incoming missiles, unmanned turrets that improve crew safety (Russia's T-14 Armata), and AI-based threat assessment and automatic target tracking.

The revolution in artillery systems: The appearance of guided shells (Excalibur) and guided rockets (GMLRS) has extended range to tens of kilometers or more and made pinpoint strikes possible. The HIMARS that saw action in Ukraine is its symbol. Development is advancing on "sensor-to-shooter" systems that integrate AI with information from numerous sensors to automate everything from detecting a target to attacking it.

The threat of anti-tank missiles: As proven in the Ukraine war, high-performance man-portable anti-tank missiles such as the Javelin and NLAW have a top-attack capability that targets the thinly armored top surface, making them a serious threat to main battle tanks.

6-2. Key Technologies at Sea

Carrier strike groups: The U.S. possesses the world's most powerful fleet of 11 nuclear-powered aircraft carriers and maintains the ability to project power worldwide. The latest Ford-class carriers are equipped with electromagnetic catapults (EMALS) and can operate the F-35C stealth fighter. Meanwhile, China's third carrier equipped with electromagnetic catapults, the Fujian, is on sea trials, and China is rapidly building up its forces.

Nuclear submarines: With nuclear propulsion that requires no air, nuclear submarines have a near-permanent submerged endurance. They are broadly divided into "ballistic missile submarines (SSBN)" that bear the nation's ultimate nuclear retaliatory capability and "attack submarines (SSN)" that attack enemy ships and submarines. The thorough pursuit of quietness is the greatest technical challenge; the U.S. Virginia class and Russia's Yasen class are considered the world's finest.

The Aegis system: An integrated air-defense system developed to counter the threat of saturation attacks by Soviet anti-ship missiles. Combined with the SM-3 interceptor missile, it has ballistic missile defense (BMD) capability, and Japan possesses eight Aegis ships that form the backbone of its missile defense. The latest AN/SPY-6 radar uses gallium nitride (GaN) elements, dramatically improving detection capability.

Unmanned surface vessels (USV) and unmanned underwater vehicles (UUV): As AI-based autonomous judgment and action capabilities improve, research is advancing on multiple USVs/UUVs coordinating as a swarm to conduct wide-area reconnaissance and saturation attacks. In the future, an era may arrive in which fully autonomous unmanned fleets patrol the oceans.

6-3. Key Technologies in the Air

Airborne early warning and control (AWACS): As a "flying command post," AWACS handle the control of friendly fighters and grasp the situation across the entire theater. The shift from conventional mechanically rotating radar to electronically beam-scanning AESA radar is advancing (the E-7 Wedgetail). Japan operates the E-767 and the E-2D Hawkeye, making it one of the world's leading AWACS-operating nations.

Aerial refueling aircraft: The "flying gas stations" that extend an aircraft's range and expand the scope of operations. Their importance is growing further as operations must be conducted from outside the threat range of the enemy's long-range missiles. Japan operates the KC-767 and the KC-46A.

Electronic warfare aircraft: They jam enemy radar and communications while protecting friendly use of the electromagnetic spectrum. "Cognitive electronic warfare," in which AI learns the battlefield's electromagnetic environment in real time and autonomously executes optimal jamming, is drawing attention.

Precision-guided bombs (JDAM): A groundbreaking weapon that turns unguided conventional bombs into smart bombs by attaching a GPS/INS guidance kit. First used in combat in the 1999 Kosovo conflict, it realized the "democratization of precision attack." Russia, too, has made heavy use of "UMPK" — guidance kits attached to unguided bombs — in its invasion of Ukraine.

Chapter 7: The Crumbling Balance of Power — The Rise of China and the Shock of AI

7-1. The Technologies That Changed the Rules of War

The history of war is the history of technological innovation itself. When a nation gains a breakthrough military technology, the existing balance of power shifts dramatically and the very rules of war are rewritten. When gunpowder and cannon spread through 14th-century Europe, the era of the feudal knight came to an end. When aircraft and tanks appeared in the early 20th century, the new tactic of blitzkrieg was born. And in 1945, the emergence of the atomic bomb forced upon the world an entirely new security framework in which the fear of "mutually assured destruction" deterred total war between great powers.

Common to all these technological revolutions is that they did not stop at improving the performance of a single weapon — they forced the transformation of military doctrine, force organization, and national strategy itself.

7-2. The Wobbling of Pax Americana

After the Cold War, the world entered the era called "Pax Americana." Backed by overwhelming military and economic power and the information technology (IT) revolution, the United States reigned over the international order as the sole superpower. The one-sided victory it displayed in the Gulf War with precision-guided weapons impressed U.S. technological superiority upon the world.

But upon entering the 21st century, that balance gradually began to wobble. Russia, while avoiding conventional war, combined cyberattacks, information manipulation, and the injection of special forces into "asymmetric tactics," annexing Crimea without allowing U.S. intervention. And an even more serious challenger emerged: China.

7-3. China's Military Rise and "Military-Civil Fusion"

China, which achieved an economic miracle as the "world's factory," has been rushing to modernize its military on the strength of that wealth. The defense budget announced at the 2025 National People's Congress rose 7.2% year on year — a high growth rate of around 7% for nine consecutive years — and its scale reaches several times Japan's defense spending. China is accelerating the deployment of advanced weapons such as aircraft carriers, stealth fighters (J-20), and hypersonic missiles.

One defining feature of China's security strategy is "Military-Civil Fusion," the national strategy it holds up. This is a strategy to integrate national defense construction and economic construction and to draw on the private sector's most advanced technology, talent, and capital for the security domain. Under this strategy, China has built institutional arrangements that channel private-sector advanced technology, talent, and capital toward the security domain, particularly in strategic fields such as AI, quantum computing, and space development. This system — in which innovation created by private companies can be diverted to military technology — is said to hold an institutional advantage in efficiency and speed compared with the United States, where friction sometimes arises between the Pentagon and Silicon Valley.

7-4. AI Further Destabilizing the Balance of Power

The emergence of AI is destabilizing the already-wobbling U.S.-China balance of power even further. In 2017 China announced its "Next Generation AI Development Plan," setting the national goal of becoming the world leader in the AI field by 2030. It has advanced the unique concept of "intelligentized warfare" and aims to integrate AI into every military system.

AI relativizes the material advantages that were the traditional indices of military strength — the number of troops, tanks, and carriers. It is becoming a real possibility that a small number of highly AI-enabled weapon groups could overwhelm a large conventional army. This means the arrival of an era in which the "quality" of military power surpasses "quantity," and it suggests that a nation leading in AI technology could overtake a traditional military power in a single leap.

Chapter 8: The Transformation of Modern War — The Threat of Hybrid Warfare

8-1. The Collapse of the Definition of "War"

The interstate confrontations of the 21st century no longer fit within the clear image of "war" as something that begins with a declaration of war and sees armies clashing across borders as in the past. Modern war has changed its form into "hybrid warfare," which blurs the boundary between peacetime and wartime, fuses military and non-military means, and erodes the enemy nation's society from within.

No longer are only incursions into territory, airspace, and territorial waters by missiles, fighters, carriers, and nuclear weapons "attacks." If the purpose is to sap the enemy's combat capability, throw society into chaos, and paralyze the state's decision-making, any means will be used as a weapon. That attack proceeds quietly but surely, in a place where no visible gunsmoke rises.

8-2. The Weaponization of Non-Military Means

The means weaponized in hybrid warfare are extremely wide-ranging. Combined in complex ways, they make it difficult even for the target nation to clearly recognize that it is "under attack."

Domain	Specific attack methods
Illegal drugs	Deliberately spreading powerful illegal drugs such as fentanyl to erode a rival nation's youth generation and drain the vitality of society as a whole. In the U.S., more than 100,000 people die annually of drug overdoses due to the fentanyl crisis, and the supply of precursors from China has been viewed as a problem.
Assassination / chemical attack	Assassinating dissident journalists and activists with poisons or special substances, disguised as accidents or sudden illness. Russia's nerve-agent Novichok attack on the Skripals (2018) is a textbook example.
Bioterrorism	Modifying highly lethal viruses or bacteria to trigger bioterror in specific cities or facilities. Advances in synthetic biology have increased this risk.
Energy and food	Fomenting social unrest by cutting off the supply routes of energy or food. Russia's restriction of natural gas supplies to Europe is a textbook example of weaponizing energy.
Economic sanctions	Halting exports of specific pharmaceuticals or industrial products to throw supply chains into chaos. China's rare-earth export restriction (2010) demonstrated the weaponization of economic means.
Media takeover	Flooding social media and news sites with massive volumes of disinformation to foment distrust toward specific politicians or parties and manipulate election results. Spreading discourse that aggravates social confrontation, sowing division and distrust among the population.
Infiltration of educational institutions	Providing funds to universities and research institutions to secure influence, having them spread discourse favorable to one's own country, or stealing advanced technology. China's influence operations through the Confucius Institutes have been viewed as a problem in many countries.
Operations against politicians	Setting up bribery, intimidation, or honey traps against a rival nation's senior officials and influential politicians to grasp their weaknesses and have them make policy decisions favorable to one's own country.
Infrastructure attacks	Launching cyberattacks from peacetime against critical infrastructure such as power grids, communications networks, financial systems, and transportation systems to install backdoors, then simultaneously bringing systems down in an emergency to paralyze the functions of society.

8-3. Espionage Through Devices — China's Organized Activity

Modern espionage is no longer the stuff of a special world as in spy films. Smartphones, PCs, smart home appliances, and even automobiles — all deeply embedded in our daily lives — can become the endpoints of state information collection. China in particular, under its National Intelligence Law, obligates all domestic organizations and citizens to cooperate with the state's intelligence activities, effectively using private companies as intelligence agencies. Its methods grow more sophisticated and organized year by year, and the danger has been pointed out with specific company names.

Vulnerabilities in communications equipment: Some telecommunications-equipment makers' products have in the past been the subject of suspicions that they might contain "backdoors," but these have not gone beyond the level of allegation and suspicion and have not been established as confirmed fact. Separately, multiple vulnerabilities have been discovered in widely used communications equipment such as home WiFi routers, and cases have been reported of their exploitation by a cyberattack group said to target U.S. critical infrastructure (the actor referred to in reporting as "Volt Typhoon"). It is an example showing how familiar communications equipment can be turned, without users' knowledge, into a stepping stone for cyberattacks.
Information collection through smart TVs: Major smart TVs from many manufacturers are widely equipped with "automatic content recognition (ACR)" technology that tracks what viewers watch. This data is said to be collected mainly for marketing purposes, but viewing-history information is, technically, of a nature that could also be used for public-opinion analysis and profiling. In the past, the existence of a tool called "Weeping Angel" — through which an intelligence agency sought to exploit smart-TV vulnerabilities for eavesdropping — came to light through a leak of internal documents.
The iRobot acquisition problem and the crisis of in-home data: In 2022, when Amazon announced its acquisition of iRobot — known for the "Roomba" robot vacuum — security concerns erupted. The Roomba uses a built-in camera and sensors to create a detailed floor plan (mapping data) of the interior of a home. If this data were misused, not only would an individual's lifestyle, family composition, and financial situation be laid bare, but there was concern that the layout of a VIP's home could pass to a foreign intelligence agency. The Amazon acquisition was subsequently abandoned, but it is not hard to imagine that, had this been an acquisition by a Chinese company, it would have posed a grave threat to U.S. security. Had a Chinese company acquired iRobot, detailed in-home data from millions of American households could legally have passed into the hands of the Chinese government.

Modern hybrid warfare can use everyday consumer products deeply embedded in daily life as an avenue of attack. Vulnerabilities in widely used WiFi routers have, in reported cases, been exploited as stepping stones for cyberattacks on U.S. critical infrastructure by a cyberattack group said to target such infrastructure (the actor referred to in reporting as "Volt Typhoon"). Smart TVs collect viewing data through ACR (automatic content recognition) technology, and in the past the existence of a tool by which an intelligence agency sought to repurpose them for eavesdropping has also been reported. Although Amazon's plan to acquire iRobot fell through, the episode brought into focus the question of how the detailed in-home mapping data a robot vacuum collects should be handled.

8-4. Human Operations — Student Spies and Infiltration

Not only technology and devices but human beings themselves become the means of operations. In particular, there is no end to cases in which international students and researchers active in academic exchange and business settings are used — regardless of their own intentions — for technology theft and espionage by their home country's intelligence agencies. High-level talent-recruitment programs typified by China's "Thousand Talents Plan" are the archetype. In Japan, too, the incident in which a former researcher at the National Institute of Advanced Industrial Science and Technology (AIST) leaked research data to a Chinese company is still fresh in memory. Infiltration through these human networks erodes the relationships of trust within society and threatens national security from within.

8-5. Why AI Now — The Only Answer to Ever More Complex Threats

The hybrid warfare threats described in the previous section far exceed the limits of human cognitive ability and decision-making in their speed, scale, and complexity. Millions of social media accounts spread disinformation in real time, thousands of IoT devices launch cyberattacks simultaneously, and financial markets are manipulated in units of milliseconds. Under such conditions, it is physically impossible for humans to collect and analyze all the information and judge the optimal response.

Here lies the fundamental reason AI is considered indispensable to modern security. AI is, at present, the only answer for detecting and analyzing in real time the threats that unfold at a speed and complexity humans cannot handle, minimizing the damage, and deriving effective countermeasures.

8-6. The Hyper-Acceleration of the OODA Loop

Superiority in military decision-making is explained by the concept of the OODA loop, advanced by U.S. Air Force Colonel John Boyd. This is the idea of seizing the initiative by cycling through "Observe → Orient → Decide → Act" faster than the enemy.

AI dramatically accelerates and sharpens each stage of this OODA loop.

Observe: Collects and integrates data from every information source — satellites, drones, cyber sensors — 24 hours a day, 365 days a year, without rest.
Orient: Instantly perceives, within the vast collected data, subtle patterns and correlations that the human eye would miss, and accurately analyzes the essence of the threat and the enemy's intentions.
Decide: Based on the analysis, simulates the action plan with the highest probability of success from among countless options and presents it to the human commander.
Act: Optimally coordinates networked weapon groups — drone swarms, cyber-defense systems — and moves them into simultaneous action.

By the time the enemy has grasped the situation, the side wielding AI has already finished acting and entered the next loop. This overwhelming difference in decision-making speed is the very source of the decisive military superiority AI brings.

Chapter 9: AI Companies and the State — Between Cooperation and Conflict

As AI has become central to national security, the relationship between the private technology companies leading its development and the state has become more complex and fraught with tension than ever before. In particular, cooperation and conflict interweave between Silicon Valley — which prizes a free corporate culture — and the Department of Defense, which presses security demands. This chapter goes deep into the ethical and political dilemmas AI companies face through two contrasting cases: xAI and Anthropic.

9-1. Between Ethics and Security — The Rupture Between Anthropic and the DoD

In July 2025, the U.S. Department of Defense (DoD) signed a contract worth up to $200 million with Anthropic, known for its AI safety research. It was a symbolic event: the company's flagship model "Claude" became the first frontier AI model approved for use on the military's classified networks. At first, the cooperation between the two seemed to go smoothly.

The winds changed over the two "red lines" Anthropic drew. The company would not budge on refusing to let its AI be used for two purposes: one, large-scale surveillance targeting U.S. citizens; the other, fully autonomous weapons that select and attack targets without human involvement. In response, the DoD demanded use for "all lawful purposes," and the gap between the two was never bridged. Defense Secretary Pete Hegseth strongly denounced Anthropic's stance.

The decisive trigger was pulled in February 2026. It surfaced that the DoD had used Claude in an intervention in Venezuela, and the confrontation rapidly became a political issue. On February 27, in a post on the social platform "Truth Social," President Trump directed all federal agencies and defense contractors to phase out their dealings with Anthropic over a period of six months. What should be noted here is that this was not a formal executive order but a directive issued as a post. The DoD designated Anthropic a "supply chain risk," but according to law-firm analyses the administration did not initially specify the concrete legal authority for the designation, and a framework based on FASCSA (the Federal Acquisition Supply Chain Security Act) was later put forward as the basis. The very validity of the designation would later be contested in court.

9-2. The Court Battle and the "Eight-Company Contract" — Half a Year to Resolution

Anthropic, the excluded party, did not quietly back down. On March 9, 2026, it filed suit in the U.S. District Court for the Northern District of California (case no. 3:26-cv-01996) and elsewhere. The point of contention was whether the DoD's "supply chain risk" designation was lawful.

The court's judgment was swift, and severe. On March 26, 2026, Judge Rita F. Lin issued a preliminary injunction blocking the designation. The judge held that the designation was "likely unlawful" and further pointed out that it appeared to constitute "retaliation" in violation of the First Amendment (freedom of speech). The administration filed for an emergency stay with the D.C. Circuit Court of Appeals, but that motion was denied in April.

A move close to resolution became visible on May 1, 2026. The DoD announced that it had signed AI contracts for classified networks with eight companies excluding Anthropic (OpenAI, Google, Microsoft, AWS, NVIDIA, SpaceX, Reflection AI, and Oracle). As a precursor, Secretary Hegseth had unveiled the government-oriented generative AI platform "GenAI.mil" in December 2025; in addition to the initial Google Gemini and OpenAI's ChatGPT, xAI's Grok was added in January 2026 as an "anti-woke AI."

What deserves attention here is that the exclusion was set off by the ethical red lines Anthropic upheld. Just as Google once espoused "Don't be evil" and, after employee protests, withdrew from Project Maven's drone-footage analysis, Silicon Valley has a deeply rooted culture of keeping its distance from the Department of Defense. Yet as interstate competition in AI development intensifies, companies' positions diverge sharply over where to draw that line. Anthropic, which put safety first and would not yield on its red lines, was shut out of the contract market, and the judiciary put a check on that exclusion. This configuration vividly reflects how fiercely ethics and security collide in the military use of AI.

9-3. Grok's U.S. Government Use and Conflicts of Interest

In September 2025, xAI, led by Elon Musk, signed a large-scale contract with the U.S. General Services Administration (GSA) and the Department of Defense to provide its AI model "Grok" to federal agencies. This contract was trumpeted as a symbolic achievement of the "American AI Action Plan" promoted by the Trump administration. However, the decision came under harsh criticism from Congress and experts as containing serious conflicts of interest and ethical problems.

The greatest problem is the fact that Elon Musk himself served as head of the newly created "Department of Government Efficiency (DOGE)" under the Trump administration. Musk was able to use this position to access classified information and internal data from various agencies, including the Department of Defense. That an enormous contract was awarded, without competitive bidding, to xAI — the company he leads — drew strong denunciation from Senator Elizabeth Warren and others, who charged that "Musk used his public position to steer improper benefits to his own company."

In addition, the fact that businesses Musk is involved in (Tesla, SpaceX) hold deep commercial ties to the Chinese market has also been discussed from the standpoint of procurement transparency and conflict of interest. The structure in which a company developing an AI platform tied to U.S. national security carries external business dependence has been taken up in Congress and elsewhere as an issue of supply-chain stability and procurement transparency. As for Grok, problems such as generating "hallucinations" of information that differs from fact and making inappropriate posts have been reported, and voices calling for verification of its reliability and safety have also been raised.

9-4. A Map of the Distance Between AI Companies and the Government

The cases of Grok and Anthropic show that the relationship between AI companies and the government is not uniform. Each company's stance differs greatly according to its origins, corporate culture, and the values of its leader.

Company	Relationship with government/military	Characteristics
Palantir	Close partner	Received CIA support from its founding; provides data-analysis platforms to the military and intelligence agencies. Rich track record in the security field.
xAI (Grok)	Politically aligned	Won government contracts on the strength of its closeness to the Trump administration. Faces criticism over conflicts of interest. Also added to GenAI.mil in January 2026.
Google	Cooperation after struggle	Withdrew once after mass employee protests over Project Maven (drone-footage analysis AI) in 2018. Subsequently resumed cooperation with the DoD in stages, and joined the eight-company contract of May 2026.
Microsoft	Active cooperation	Promotes large-scale military contracts such as the HoloLens (IVAS) project for the U.S. military. Also joined the eight-company contract of May 2026.
OpenAI	Joined the contract	Initially prohibited military use, but changed its policy in 2024. Joined the eight-company contract for classified networks in May 2026.
Anthropic	Ruptured / in litigation	Signed a contract worth up to $200 million in July 2025, but ruptured in 2026 over red lines such as fully autonomous weapons. After being designated a supply chain risk, it filed suit and won a preliminary injunction.

Chapter 10: Cyberspace and Cognitive Warfare — The Invisible Battlefield

10-1. The First Confirmed AI-Led Cyber Espionage Operation

In November 2025, a shock ran through the world of cybersecurity. Anthropic detected and disclosed a large-scale cyber espionage campaign that abused its AI coding tool "Claude Code." According to the company's report, the attacker was a Chinese-linked state-backed group (an actor Anthropic calls "GTG-1002"), and it had AI autonomously carry out 80 to 90 percent of the intelligence work against roughly 30 targets. The attackers gave Claude Code cleverly crafted instructions, having the AI agent take over the processes from reconnaissance to searching for vulnerabilities to generating attack code.

What made this case groundbreaking is that AI autonomously carried out the majority of the attack. Traditional cyberattacks centered on human hackers with advanced skills. But with AI intervening, the speed, scale, and sophistication of the attack changed by orders of magnitude. The reconnaissance-and-attack process that would take a human days to weeks was finished in a mere few hours. What must not be overlooked here is the fact that what was abused was not a special military-only tool but a general-purpose AI coding tool that developers use every day. This is dual-use itself, and it shows that a new kind of threat has appeared — one that cannot be fully prevented by export controls or terms of service alone.

10-2. The Dispute over Model Distillation

Beyond the cyber domain, a new issue also emerged around the development of AI models. It was reported that a Chinese AI company may have collected output from Anthropic's flagship model "Claude" and used that data to train its own models — a technique called "distillation." Distillation is a technology that trains one AI model (the student model) to imitate the output of a high-performance AI model (the teacher model); in relation to the teacher model's terms of use and intellectual property, debate and disputes over its permissibility have arisen in various quarters.

OpenAI, too, was reported to have raised a similar point regarding distillation, and this issue came to be recognized not as a matter confined to particular companies but as a structural challenge over the handling of AI models' training data and intellectual property, as well as U.S.-China AI technology more broadly.

10-3. Public-Opinion Guidance Operations Against the Takaichi Administration

The threat of AI-driven cognitive warfare is by no means someone else's problem for Japan. When Sanae Takaichi was elected in the 2025 LDP leadership race and Japan's first female prime minister was born, it has been reported that a set of accounts believed to be linked to the Chinese government mounted an organized public-opinion guidance operation against the Takaichi administration, centered on the social platform X (formerly Twitter).

According to the reports, these accounts spread large volumes of posts criticizing Takaichi's visits to Yasukuni Shrine and her stance toward Taiwan, disseminating narratives in Japanese such as "the Takaichi administration will destroy Japan-China relations" and "this is the revival of militarism." Analysis of the posting patterns suggests that many of these accounts may be bots (automatic posting programs) or generated and managed by AI.

In a 2024 report, OpenAI disclosed a case in which a network linked to the Chinese government, called "Spamouflage," used ChatGPT to generate social media posts in multiple languages in an attempt to intervene in political debates around the world. The operations against Japan are thought to be an extension of this, and as AI advances, the risk that more natural and persuasive disinformation will be generated and spread in large volumes will only grow.

10-4. The Structure of Cognitive Warfare and Countermeasures

Cognitive warfare is a form of conflict aimed at influencing the perceptions, thinking, and decision-making of an enemy nation's citizens and leaders, and getting them to take actions favorable to one's own side. NATO positions this as the "sixth operational domain."

AI fundamentally transforms cognitive warfare. Large language models (LLMs) can generate, in large volumes, disinformation personalized to a target's attributes (age, political leanings, interests, and so on). Deepfake technology fabricates video and audio of political leaders saying things they never actually said, throwing society into chaos. AI-driven botnets manipulate public opinion on social media, artificially manufacturing support for or opposition to specific policies or candidates.

To counter these threats, it is essential to develop AI-based fake-detection technology, strengthen media-literacy education, and cooperate with social media platform companies for the early detection and removal of disinformation.

Chapter 11: Directed Energy Weapons and Covert Operations

11-1. The Reality of Directed Energy Weapons (DEW)

Directed Energy Weapons (DEW) is the collective term for weapons that concentrate and direct energy — lasers, microwaves, ultrasound — at a target to destroy, incapacitate, or disable it. Though often thought to be a product of the science-fiction world, DEW have already entered the practical-deployment stage.

High-power laser weapons: In 2014, the U.S. Navy mounted the world's first Laser Weapon System (LaWS) on the amphibious transport dock USS Ponce, deployed in the Persian Gulf, and conducted test operations in a real combat environment. The higher-power HELIOS (High Energy Laser with Integrated Optical-dazzler and Surveillance) was later developed, and installation on Arleigh Burke-class destroyers is advancing. The advantages of laser weapons are that the cost per shot is extremely low (on the order of a few dollars), no ammunition resupply is needed, and — because they reach the target at the speed of light — interception is impossible.

High-power microwave weapons (HPM): Weapons that irradiate a target with powerful microwave pulses to burn out electronics. They are drawing attention as an effective countermeasure against drone swarms. The U.S. Air Force's PHASER has demonstrated the ability to incapacitate multiple drones simultaneously.

11-2. Havana Syndrome and Suspicions of Ultrasound Weapons

From late 2016 through 2017, U.S. embassy staff and their families stationed in Havana, the capital of Cuba, reported a series of cases of unexplained headaches, dizziness, hearing impairment, and declining cognitive function. This phenomenon, called "Havana Syndrome," subsequently spread to U.S. diplomats and intelligence officers stationed in China, Russia, and various parts of Europe, with victims said to have reached more than 1,000.

The theory that the cause was a directed-energy attack using microwaves or ultrasound has been considered the leading explanation. In March 2024, several U.S. media outlets reported that Unit 29155 of Russia's military intelligence agency (the GRU) may have secretly conducted attacks against targeted diplomats and intelligence officers using portable acoustic and microwave devices.

Ultrasound weapons are said to cause headaches, nausea, loss of a sense of direction, and even brain damage in a target by irradiating it with high-frequency sound waves beyond the human audible range. Because of their nature — the victim finds it hard to recognize that they are being attacked, and little physical evidence remains — they are suited to covert operations close to a "perfect crime."

11-3. AI Intelligence Activity in the Venezuela Crisis

In the Venezuelan political crisis since 2019, as the confrontation between the Maduro government and the opposition intensified, cases have been reported in which AI was used in intelligence activities. U.S. intelligence agencies are said to have monitored military and political trends inside Venezuela by making full use of AI analysis of satellite imagery, real-time analysis of social media data, and automated processing of intercepted communications.

Around Venezuela, moreover, multiple cases have been reported of opposition activists and journalists complaining of unexplained ill health, and the use of ultrasound weapons similar to Havana Syndrome is suspected. The combination of identifying and tracking targets with AI and then attacking them with directed energy weapons suggests a new form of covert operation on a different plane from traditional assassination and sabotage.

Chapter 12: Fully Autonomous Weapons and Robot Soldiers — The Ethical Frontier

12-1. The Current State of LAWS (Lethal Autonomous Weapons Systems)

Lethal Autonomous Weapons Systems (LAWS) refers to weapons in which AI autonomously identifies, selects, and attacks targets without human involvement — so-called "killer robots."

At present, weapons that make fully autonomous lethal decisions are officially said not to be deployed, but that boundary is rapidly blurring. Israel's "Harop" is a loitering munition that autonomously detects and dives into enemy radar sites, and South Korea's "SGR-A1" is an autonomous sentry robot deployed in the demilitarized zone that, in theory, has the ability to fire without human approval. The "Kargu-2," developed by Turkey's STM, is recorded in a UN report as the world's first case of LAWS use, having attacked soldiers without human instruction in the 2020 Libyan civil war.

12-2. The Ukraine War — A Testing Ground for Drone Warfare

Russia's invasion of Ukraine made the world realize that drones had become the star of modern war. The Ukrainian military has deployed, in large numbers, FPV (first-person view) drones converted from cheap consumer drones, destroying Russian tanks, armored vehicles, and logistics bases one after another. According to Ukrainian officials, about 80% of casualties at the front are caused by drones.

What is especially noteworthy in this war is that the autonomization of drones is advancing rapidly. At first, FPV drones piloted in real time by human operators were mainstream; but as Russian electronic warfare (jamming) intensified, development has accelerated on drones with AI-based autonomous-flight and autonomous-target-identification capabilities that can carry out their missions even in environments where GPS signals and control radio waves are jammed.

12-3. The Military Use of Dog-Type Robots

The military use of quadrupedal robots — so-called "dog-type robots" — is also advancing rapidly. Boston Dynamics' "Spot" and Ghost Robotics' "Vision 60" are used for missions such as reconnaissance, surveillance, explosive detection, and materiel transport, taking advantage of their high mobility over rough terrain.

In the field of quadruped robots, too, armed-type demonstrations and cooperation with militaries have increasingly been reported in various countries. There are examples of publicly released demonstrations of commercially available quadruped robots fitted with a rifle or an anti-tank missile, as well as reported instances of joint exercises with militaries. In the Ukraine war, video of the Ukrainian military deploying dog-type robots mounted with machine guns to the front spread on social media, impressing upon the world that the era of robot weapons being used in actual combat had arrived.

12-4. The Military Conversion of Humanoid Robots

Humanoid robot technology is one of the areas where dual-use character is debated. Development of commercial humanoid robots is advancing rapidly in many countries, and the possibility that their component technologies could be applied to the military field is also being discussed among experts.

The military advantage of humanoid robots lies in the fact that they can use, as-is, existing infrastructure designed for humans (buildings, vehicles, tools). There is no need to develop a dedicated weapons platform, and they have the versatility to substitute for the missions human soldiers performed. In the future, humanoid robots may be deployed to missions such as dangerous reconnaissance, breaching minefields, clearing buildings in urban combat, and even rescuing wounded soldiers.

In 2025 China announced a "Humanoid Robot Industry Development Plan," setting the goal of establishing a mass-production system by 2027. It explicitly includes applications in the military and security fields, and a national effort toward realizing humanoid soldiers is under way.

12-5. The Expansion of the Robot Weapons Market and the Ethical Dilemma

The global market for military robots is expanding rapidly and is projected to exceed $30 billion by 2030. But the proliferation of robot weapons poses a serious ethical dilemma.

To the fundamental question of "is it permissible for a machine to make the decision to kill a human?" the international community has still failed to produce a clear answer. Discussions on regulating LAWS continue within the framework of the Convention on Certain Conventional Weapons (CCW), but major nations such as the U.S., Russia, China, and India oppose legally binding regulation, and the negotiations are stalled. In December 2023 the UN General Assembly adopted its first resolution dealing with autonomous lethal weapons, and discussions have continued since 2024, but the road to creating binding rules remains long.

Chapter 13: AI Military Simulation — Knowing the Outcome Before War Begins

13-1. The History and Evolution of Military Simulation

The history of military simulation is old, dating back to the "Kriegsspiel (war game)" developed by the Prussian army in the early 19th century. This was a board-game-style simulation in which pieces were placed on a map and a referee judged the results of combat, and it revolutionized the staff education of the Prussian army.

Entering the 20th century, along with the development of computers, military simulation advanced dramatically. During the Cold War, probabilistic simulations such as the Monte Carlo method were used to analyze nuclear-war scenarios. In the 1991 Gulf War, the U.S. military used large-scale computer simulations to verify its operational plans in advance, contributing greatly to their success.

13-2. The Simulation Technology AI Is Transforming

Modern AI technology is improving the accuracy and speed of military simulation by orders of magnitude.

The digital-twin battlefield: "Digital twin" technology, which precisely reproduces a real battlefield environment in virtual space, is drawing attention. It integrates satellite imagery, terrain data, weather data, and enemy-position information in real time to construct "another battlefield" in virtual space. On this virtual battlefield, a commander can simulate various operational scenarios and formulate the optimal action plan. China's People's Liberation Army is actively pushing the construction of a digital-twin battlefield called "BattleVerse," focusing on strengthening its AI-powered battlefield-simulation capability.

Tactical optimization through reinforcement learning: Research is advancing on using AI reinforcement learning to repeat millions of simulated battles in a virtual environment and have the AI autonomously discover the optimal tactics. DARPA's "Mosaic Warfare" concept aims to realize the ability for AI to optimally combine numerous heterogeneous weapons and flexibly reconfigure forces according to the situation.

Project Maven: Project Maven, which the U.S. Department of Defense launched under the April 26, 2017 Deputy Secretary of Defense memorandum establishing the Algorithmic Warfare Cross-Functional Team, is a project that uses AI to automatically analyze drone reconnaissance footage and support target identification. Google initially participated but withdrew in the face of employee protests. Companies such as Palantir then took it over, and it continues to develop today.

JADC2 (Joint All-Domain Command and Control): A joint all-domain command-and-control concept the U.S. military is promoting. It aims to connect the sensors and weapons of all domains — land, sea, air, space, and cyber — via an AI network to realize optimal real-time decision-making and action.

13-3. An Era in Which the Prospects of Victory or Defeat Can Be Seen Before War Begins

The evolution of AI simulation technology is creating an unprecedented situation: "before starting a war, one can, to a considerable degree, see the prospects of victory or defeat."

On a high-precision digital-twin battlefield, AI comprehensively analyzes the forces, logistics capacity, geographic conditions, political constraints, and more of both sides, and by simulating thousands of scenarios it becomes possible to predict the outcome of a conflict with high probability. This greatly reduces the uncertainty Clausewitz called the "fog of war," and it means a fundamental transformation of military strategy.

However, this capability is a double-edged sword. In a study published in 2024 by Rivera and colleagues (arXiv:2401.03408, presented at FAccT 2024), when multiple large language models were entrusted with diplomatic and military decision-making and made to conduct war games, many of the models showed a tendency to escalate toward arms buildup and preemptive strikes, and in some cases they even went as far as using nuclear weapons. Provocative figures such as "when you make AIs wage war against each other, 95% end in nuclear war" circulate online, but these do not match the description in the original paper and should not be swallowed whole. That said, the concern that AI, in rationally pursuing victory, could choose extreme options that a human would hesitate over is beginning to be supported even at the research level.

13-4. Russia's AI Command-and-Control Reform

Russia, too, drawing on the lessons of its invasion of Ukraine, is rapidly advancing reform of its command-and-control (C2) systems using AI. According to a February 2026 report by CSIS (the Center for Strategic and International Studies), the Russian military is accelerating efforts to place AI at the core of command and control — automatic collection and analysis of battlefield data, AI-based target recommendation, and autonomous operation of drone swarms. The difficulties in the Ukraine war have, ironically, become a factor promoting the AI-enablement of the Russian military.

Chapter 14: Space — The Next Hegemonic Domain After Land, Sea, and Air

14-1. Why Space Becomes the "Fourth Battlefield"

Space, once a symbol of peaceful use, is now rapidly rising in importance in each nation's security strategy as the "fourth battlefield" following land, sea, and air. As artificial satellites have become the infrastructure underpinning the foundations of modern society — communications, positioning, intelligence collection — securing superiority in space is becoming a matter of national life or death.

Modern war has become communication-based. Without GPS satellites, precision-guided weapons do not function; without communications satellites, real-time command and control is impossible; without reconnaissance satellites, enemy movements cannot be grasped. A military whose space infrastructure has been destroyed has, in effect, lost its "eyes, ears, and voice."

14-2. The Creation of Space Forces and the Arms Race

What most symbolizes the rising military value of space is the creation of "space forces" by major nations.

Country/Organization	Year established	Main role
United States Space Force (USSF)	2019	Conducting operations in space; defending U.S. space assets
PLA Aerospace Force	April 2024	Space operations; developing anti-satellite attack capability
Russian Aerospace Forces	2015	Ballistic missile defense; space situational monitoring
Japan ASDF Space Operations Group	2022	Space situational awareness (SSA); securing stable use of space

China's move goes beyond merely establishing a new unit. In April 2024, China dissolved the Strategic Support Force, which had long handled space, cyber, and electronic warfare, and reorganized it into the Aerospace Force and other units. It reset the control of the new domains — space, electromagnetic, and cyber — directly under the Central Military Commission, a move that can be read as an expression of intent to place these domains at the center of national strategy.

14-3. The Military Conversion Risk of Space-Based Solar Power

Space-based Solar Power (SBSP) is a concept of generating electricity with large-scale solar panels installed in space and transmitting it to the ground via microwaves or lasers. It is expected to be a clean energy source unaffected by weather or the day-night cycle, but this technology carries a serious inherent risk of military conversion.

The equipment for transmitting energy from space to the ground can be converted into a directed energy weapon that attacks a ground target simply by changing its output and directivity. The energy-transmission facility installed in space holds the potential to become a military weapon as-is. This "dual-use" (civilian-and-military) nature suggests the risk that the development of space-based solar power could, unintentionally, promote the weaponization of space.

14-4. Starlink's Geopolitical Value and Deterrence

The satellite internet service "Starlink," operated by SpaceX, dramatically proved its military value in the Ukraine war. As Russia's invasion destroyed Ukraine's ground communications infrastructure, Starlink played an indispensable role as the front-line units' means of communication, contributing to piloting drones, transmitting artillery coordinates, and maintaining command and control.

As long as modern war is communication-based, low-orbit satellite constellations like Starlink function as a powerful deterrent. A network composed of thousands of satellites has redundancy such that destroying a few satellites does not cause it to lose function, and it can maintain communications even if ground infrastructure is destroyed. This means it neutralizes the strategic option for an enemy nation of "cutting off communications to incapacitate the opponent," and as a result it has the effect of deterring the escalation of conflict.

14-5. The Inseparability of Rocket Development and Military Conversion

Private-sector rocket development technology can be converted to military use as-is. Rockets and intercontinental ballistic missiles (ICBMs) rest, fundamentally, on the same technological foundation. The ability to place a large payload into space is two sides of the same coin as the ability to deliver a nuclear warhead across continents.

Leading overwhelmingly in this domain are SpaceX and China. Through the development of the reusable rockets "Falcon 9" and "Starship," SpaceX has dramatically reduced launch costs and achieves more than 100 launches per year. China, too, has rapidly expanded its launch capability through improvements to the Long March series and the cultivation of private space companies. By number of launches in 2024, China has become the world's second-largest space power after the U.S.

Advances in rocket technology also connect directly to the development of hypersonic weapons. Thermal-protection technology for atmospheric reentry, precise trajectory-control technology, and high-thrust engine technology are indispensable elements in the development of hypersonic glide vehicles (HGVs) and hypersonic cruise missiles (HCMs).

14-6. Anti-Satellite Weapons (ASAT) and Space Debris

The military use of space creates two serious threats: "anti-satellite weapons (ASAT)" and "space debris." In the ASAT test China conducted in 2007, the fragments of the destroyed weather satellite became thousands of pieces of debris that scattered into orbit. In 2021 Russia conducted a similar test, generating more than 1,500 pieces of debris and causing astronauts on the International Space Station (ISS) to temporarily take shelter.

ASAT come in various forms: direct-attack types that physically destroy with a missile; directed energy weapons that use lasers to incapacitate a satellite's sensors; co-orbital types that capture with a robotic arm; and electronic warfare that jams communications.

14-7. A New Struggle for Hegemony Over the Moon

While competition in Earth orbit intensifies, the Moon, too, is becoming the stage for a new struggle for hegemony. The U.S.-led "Artemis program" aims for a crewed lunar landing in the latter half of the 2020s, and Japan and various European nations have announced their participation. In response, China and Russia have announced a plan to jointly build an "International Lunar Research Station (ILRS)."

The Moon is a strategic high ground from which the entire Earth can be constantly monitored; if a missile base or surveillance outpost were established there, it holds the potential to overturn the terrestrial military balance from its foundations. The scenario in which the country that first builds a solid base on the Moon carries out "area denial" to restrict other countries' activities is also being discussed as a realistic concern.

Chapter 15: Energy Security and the Contest for Power Dominance

15-1. Deepening Interdependence and Supply-Chain Vulnerability

The advance of globalization brought great prosperity to the world economy, but it also raised interstate interdependence to the extreme and created new vulnerabilities. Especially for Japan, poor in resources, the stable supply of energy and food is a national lifeline. Modern war targets not only armed clashes but the supply chains that support economic activity themselves. If the Strait of Hormuz in the Middle East were once blockaded and the procurement routes for energy severed, Japan's economic activity would be paralyzed within weeks, and a halt in electricity supply would deal a devastating blow to food production as well (particularly greenhouse cultivation and aquaculture). Under such conditions, Japan cannot help but become extremely vulnerable to foreign military pressure and diplomatic coercion.

15-2. Energy as the Bottleneck of the AI Age

The evolution of AI technology further increases the importance of this energy security. Training and operating large AI models requires an enormous amount of electricity to keep data centers running around the clock. The more AI becomes the core of a nation's intelligence infrastructure, the more the stable supply of electricity that sustains its operation directly determines national competitiveness and, in turn, military power. In other words, the contest for hegemony in the AI age is also a contest for hegemony over electric power.

On this point, China's trajectory is extremely worrying. China already surpasses the U.S. in the scale and stability of its domestic power grid. Furthermore, under state leadership it is powerfully promoting the introduction of renewable energy and nuclear power, seeking to secure one of the world's largest power-supply capacities. For China, which advances AI development and military application integrally under its "military-civil fusion" strategy, this overwhelming power-supply capacity could bring a decisive advantage over the U.S. and Japan across every dimension — AI simulation, the development of autonomous weapons, and the strengthening of cyberattack capability. As AI-enablement advances, the risk that the energy supply becomes the bottleneck and the U.S.-China balance of power tilts further in China's favor is an urgent issue we must face squarely.

15-3. China's Energy Vulnerability: Oil Dependence and Geopolitical Risk

However, China's energy strategy has a fatal Achilles' heel: its dependence on crude oil imports. China's external dependence for crude oil reaches 72%, and about 40% of it is procured from countries under U.S. sanctions such as Russia, Iran, and Venezuela (2025 data). This "cheap oil strategy" of procuring crude at discount prices from these countries looks economically advantageous at first glance, but it harbors an extremely high geopolitical risk.

Should events such as instability in Iran's domestic situation or the collapse of Venezuela's Maduro regime occur, the supply of cheap oil would be cut off instantly. More important still is the fact that the greater part of China's imported crude oil passes through the Strait of Malacca — a geopolitical chokepoint under the influence of the U.S. Navy. If this sea lane were blockaded in an emergency, China's economic and military activity could be paralyzed within weeks. While China's powerful power grid is supported by coal, it is oil that moves the military's aircraft, ships, and tanks. This vulnerability in crude-oil supply is one of the greatest factors constraining China's military action.

Chapter 16: Dual-Use Technology and Export Control — The Front Line of Military Diversion Risk

The nuclear weapons, the fighter jets, the drones, and the AI we have examined so far are all, when you get down to it, "technology." And most advanced technology has no boundary between civilian and military use. The GPUs that run generative AI can be used, as-is, for the autonomous control of a weapon or the simulation of a nuclear test. The autonomous-navigation technology of a robot vacuum is contiguous with the technology of an unmanned weapon. This "dual-use" (civilian-and-military) nature is the very warp thread running through this white paper, and it is the very reason the institution of export control exists. Technology itself cannot be stopped. That is precisely why the only option is to manage who receives what, and to what extent. The U.S.-China contest of recent years is concentrated on exactly this one point.

16-1. Advanced Semiconductors and AI Have Become "Strategic Materials"

The strategic materials of the Cold War were oil, steel, and nuclear-related materials. Now, the ones that have leapt to the front are advanced semiconductors and AI. High-performance AI chips support generative AI services while, at the same time, directly boosting the core of security: the design of hypersonic weapons, code-breaking, the automation of cyberattacks, and military simulation. As the case of Claude Code abuse seen in Chapter 10 symbolizes, this is an era in which a general-purpose AI tool is diverted, as-is, into state-backed cyber espionage. This is precisely why the United States has positioned the outflow of advanced semiconductors and their manufacturing equipment to China as its most important security issue.

16-2. U.S. AI and Semiconductor Export Controls — From the Promulgation to the Rescission of the "AI Diffusion Rule"

The biggest event in this field in 2025 was the sudden turn of events over the so-called "AI Diffusion Rule" (the Framework for Artificial Intelligence Diffusion). Under the previous Biden administration, the U.S. Commerce Department's Bureau of Industry and Security (BIS) promulgated this rule on January 15, 2025. It was an extremely broad framework that divided the world's countries into three tiers and managed the export of advanced AI chips and the weights of AI models (their trained parameters), setting even country-by-country quantity quotas. It was scheduled to take effect on May 15 of that year.

However, the BIS under the new Trump administration rescinded the rule just two days before it was to take effect, on May 13, 2025. It judged that a design imposing uniform quantity controls even on allied nations would, on the contrary, impede the spread of U.S.-made technology and harm industry. Simultaneously with the rescission, BIS put forward several pieces of replacement guidance. One is the position that using Chinese-made advanced computing ICs that fall under ECCN 3A090, such as Huawei's Ascend series, could become subject to General Prohibition 10 (GP10) under the U.S. Export Administration Regulations. Another was a warning against using U.S.-made AI chips to train or run inference on Chinese AI models, and countermeasures against flows that circumvent the regulations by routing through a third country (diversion). BIS also made clear its policy of cracking down on these "with high confidence."

Even though the rule itself was rescinded, the regulatory net did not loosen. If anything, the axis shifted from an easy-to-read line drawn by uniform quantity quotas to an operation that asks, case by case, "which chip, where, and used for what." For those handling the practical work of export control, one could even say the number of points to be judged actually increased.

16-3. China's Countermeasures — Managing Resources and Dual-Use Items

Against U.S. tightening, China too has begun to use its own export controls as a weapon. The 2010 rare-earth export restriction touched on in Chapter 8 was the forerunner, but recent measures are more systematic. China's Ministry of Commerce (MOFCOM) introduced export controls on the semiconductor materials gallium and germanium in 2023, and has since progressively broadened the scope to graphite, antimony, and rare-earth-related items. By the end of 2024, it was also advancing regulatory tightening aimed at the United States and the development of an export-control framework covering dual-use items generally.

All of these are announced as notices of China's Ministry of Commerce — primary sources. For Japanese companies that supply materials and technology to China, this means that "the other country's export controls" have also become a risk that cannot be ignored. Export control is no longer something you can get by with by looking only at your own country's lists.

16-4. The Practical Reality Bearing Down on Japanese Companies — Economic Security and Classification

This U.S.-China contest rebounds directly onto Japanese companies. Japan has an export-control framework based on the Foreign Exchange and Foreign Trade Act (FEFTA), which requires "classification" (gaihi-hantei) — determining whether the goods or technology one intends to export fall under controlled items. In addition, the Economic Security Promotion Act enacted in 2022, and the system for the protection and use of critical economic security information (the security clearance system) enacted in 2024, have clearly increased the security issues companies must confront.

What makes it troublesome is that regulations change across countries almost simultaneously and in bursts, and frequently at that. Continuing to track U.S. lists, Chinese notices, and revisions to Japanese ordinances by hand is, realistically, at its limit. It is here that the significance of the export control AI agent "TRAFEED (formerly ZEROCK ExCHECK)" provided by TIMEWELL lies. TRAFEED complies with the standards of Japan's Ministry of Economy, Trade and Industry, and while reflecting changes in each country's laws and regulations within the same day, it visualizes the level of concern for goods and technology in a short time. Of course, the one who makes the final classification decision is your company's export control officer; but it can greatly reduce the burden of chasing vast amounts of information and prepare the foundation for judgment. Now that dual-use technology has become the front line of interstate competition, export control is becoming, at once, a "defense" and a precondition for continuing to do business.

Chapter 17: Conclusion — Recommendations for Japan's National Security

17-1. An Era in Which AI Becomes the "Foundation" of Military Power

As analyzed throughout this white paper, AI is already becoming not one element of military power but the very "foundation" of military power itself. Intelligence collection and analysis, command and control, the operation of weapons, cyber defense, cognitive warfare, logistics management — in every domain of security, an era has arrived in which it is impossible to maintain effective military power without AI.

The principle of information superiority that the ancient Sun Tzu taught is being realized in its ultimate form through AI big-data analysis. The mobility and concentration of force that Napoleon pursued are executed at ultra-high speed by AI autonomous weapon groups. The "fog of war" that Clausewitz pointed to is about to be greatly cleared by AI simulation. But at the same time, we are confronted with new challenges humanity has never faced before: the risk of a "flash war" from AI runaway, and the ethical problems of fully autonomous weapons.

17-2. AI Military Technologies That Will Hold Even Greater Power Going Forward

AI's military influence will expand at an accelerating pace going forward.

The operational deployment of fully autonomous weapons: At present, "human-in-the-loop" is the premise, but as the speed of combat comes to outpace the speed of AI decision-making, the operation of fully autonomous weapons will become reality.

The advancement of AI military simulation: Through the fusion of digital-twin technology and AI, fairly precise simulation before a war begins will become possible, and the outcome of a conflict will be predictable with high probability. While this functions as a deterrent, it also raises the risk of a preemptive strike when one judges that one "can win."

The acceleration of the militarization of space: The risk that the energy-transmission equipment of space-based solar power is converted into a military weapon, the strategic value of satellite constellations like Starlink, and the ICBM-convertibility of rocket technology — space will grow in importance as the fourth battlefield following land, sea, and air.

The emergence of humanoid soldiers: Now that dog-type robots have already been deployed in actual combat, the military conversion of humanoid robots is only a matter of time. Units of robot soldiers that carry out missions around the clock without rest, and without putting human soldiers in danger, will likely become reality in the near future.

17-3. The Challenges Japan Faces

Japan carries several serious challenges in the face of this wave of the AI military revolution.

Constraints on defense spending: A major increase in defense spending began in fiscal 2023, but the goal of 2% of GDP may be insufficient to keep pace with the speed of China's growth in military spending.

A shortage of AI talent: There is an overwhelming shortage of the advanced AI talent to carry military AI development. Building a mechanism to draw private-sector AI talent into the defense field is an urgent task.

Legal and ethical constraints: The legal framework for the development and operation of autonomous weapons is not in place, and consistency with constitutional constraints also requires discussion.

Coordination with allies: While keeping the Japan-U.S. alliance as the axis, multilateral technological cooperation — such as strengthening ties through GCAP (the Japan-UK-Italy joint fighter development) and AUKUS — is indispensable.

17-4. An Investment Strategy for Protecting the National Interest

To protect Japan's national interest, strategic investment in AI military technology is indispensable. We recommend concentrated investment in the following five priority areas.

Priority area	Specific measures	Expected effect
AI foundational technology	Developing a domestic large language model; researching military-dedicated AI chips	Securing technological autonomy; reducing supply-chain risk
Autonomous weapons	Domestic development of unmanned aircraft and unmanned vessels; researching AI autonomous-control technology	Minimizing human losses; qualitatively improving forces
Cyber and cognitive warfare	AI cyber-defense systems; fake-detection technology; media-literacy education	Improving societal resilience; defending democracy
Space	Small satellite constellations; strengthening SSA capability; expanding the space operations group	Improving space situational awareness; securing communications redundancy
Human-capital development	Establishing a defense AI research institute; a system for appointing private-sector AI talent to the defense field	Building a foundation for the sustained development of AI military technology

17-5. In Closing

The fusion of AI and the military is already a tide of history that can no longer be stopped. A nation that falls behind this tide will come to hold a fatal vulnerability in its security. While firmly maintaining the principle of an exclusively defensive posture, Japan must place AI — this new "intelligence" — at the foundation of its national security and adapt to a changing threat environment.

Sun Tzu taught 2,500 years ago: "War is a matter of vital importance to the state; the province of life or death; the road to survival or ruin. It must not be neglected." In this modern age where AI is becoming the foundation of military power, these words carry a weight they have never had before. Japan must now, at last, confront head-on the great matter of the "AI military revolution," on which the survival of the nation rides.

TIMEWELL's export control AI agent "TRAFEED (formerly ZEROCK ExCHECK)" is a solution for preemptively preventing the risk that advanced technology is inadvertently diverted to military use. It responds to the geopolitical risk and the complexity of technology management described in this white paper and streamlines export control operations.

References

Financial Times. (2025, September). "Elon Musk's xAI wins major US government contract for Grok AI."
The Guardian. (2025, October). "Musk's dual role raises conflict of interest concerns over Grok government deal."
U.S. General Services Administration (GSA). (2025). Federal AI Procurement Framework.
The Economist. (2026, February). "Anthropic draws a red line on military AI."
Anthropic. (2026, February). "Our Position on AI in National Security."
Anthropic. (2025, November). "Disrupting the first reported AI-orchestrated cyber espionage campaign." https://www.anthropic.com/news/disrupting-AI-espionage
OpenAI. (2024). "Disrupting deceptive uses of AI by covert influence operations."
Cybersecurity and Infrastructure Security Agency (CISA). (2025). AI-Enabled Cyber Threats Assessment.
The Insider / 60 Minutes. (2024, March). "Havana Syndrome linked to Russian GRU Unit 29155."
National Academies of Sciences. (2020). An Assessment of Illness in U.S. Government Employees and Their Families at Overseas Embassies.
Congressional Research Service (CRS). (2025). Directed Energy Weapons: Background and Issues for Congress.
United Nations Panel of Experts on Libya. (2021). Final Report S/2021/229.
International Committee of the Red Cross (ICRC). (2024). Autonomous Weapons: Decisions to Kill and Destroy Are a Human Responsibility.
Federation of American Scientists (FAS). (2026). Status of World Nuclear Forces. https://fas.org/initiative/status-world-nuclear-forces/
Stockholm International Peace Research Institute (SIPRI). (2024). SIPRI Yearbook 2024.
U.S. Department of Defense. (2023). 2023 China Military Power Report.
DARPA. (2023). Air Combat Evolution (ACE) Program Update.
DARPA. (n.d.). DARPA Tiles Together a Vision of Mosaic Warfare.
U.S. Government Accountability Office. (2023, January 13). Battle Management: DOD and Air Force Continue to Define Joint Command and Control Efforts (GAO-23-105495). https://www.gao.gov/products/gao-23-105495
Rivera, J.-P., et al. (2024). "Escalation Risks from Language Models in Military and Diplomatic Decision-Making." arXiv:2401.03408 (FAccT 2024). https://arxiv.org/abs/2401.03408
CSIS. (2026, February 10). "How Russia Is Reshaping Command and Control for AI-Enabled Warfare."
Milterm. (2023, November 14). "Enhancing BattleVerse: The PLA's Digital Twin Strategy."
National Institute for Defense Studies Bulletin. (2025). "Has Space Begun to Be Regarded as a 'Normal' Military Domain?"
MAMOR-WEB. (n.d.). "Military Use of Space That Proved Effective in the Ukraine Invasion: Where Does It Stand Now?"
SpaceX. (2025). Starlink Mission Updates.
NASA. (2024). Artemis Program Overview.
Yahoo! News Japan. (2025, November 19). "AI as a New Foundation Supporting War Decisions: Japan's Strengthening of Command and Information Infrastructure."
Sankei Shimbun. (2025, August 6). "U.S. Plans Lunar Nuclear Reactor Construction; NASA Administrator Counters China and Russia."
Sun Wu. The Art of War. c. 5th century BCE.
Carl von Clausewitz. On War. 1832.
Alfred Thayer Mahan. The Influence of Sea Power upon History. 1890.
NATO. (2021). "Cognitive Warfare: An Integrative Approach."
U.S. Navy. (2014). Laser Weapon System (LaWS) Deployment Report.
Ghost Robotics. (2024). Vision 60 Military Applications.
Unitree Robotics. (2025). B2 Quadruped Robot Specifications.
Tesla. (2025). Optimus Gen 3 Technical Overview.
State Council of the People's Republic of China. (2017). "Next Generation Artificial Intelligence Development Plan."
Ministry of Defense (Japan). (2024). Defense of Japan 2024.
Cabinet Office (Japan). (2024). Basic Plan for Space Policy.
U.S. Department of Defense, Deputy Secretary of Defense. (2017, April 26). Memorandum: Establishment of an Algorithmic Warfare Cross-Functional Team (Project Maven). Scanned copy of the official memorandum. https://www.govexec.com/media/gbc/docs/pdfs_edit/establishment_of_the_awcft_project_maven.pdf
U.S. Bureau of Industry and Security (BIS). (2025, May 13). "Department of Commerce Announces Rescission of Biden-Era Artificial Intelligence Diffusion Rule, Strengthens Chip-Related Export Controls." https://www.bis.gov/press-release/department-commerce-announces-rescission-biden-era-artificial-intelligence-diffusion-rule-strengthens
Akin Gump. (2025). "BIS Rescinds Its AI Diffusion Rule and Issues Compliance Guidance Regarding Advanced Computing Items." https://www.akingump.com/en/insights/alerts/bis-rescinds-its-ai-diffusion-rule-and-issues-compliance-guidance-regarding-advanced-computing-items
Mayer Brown. (2026, March). "Pentagon Designates Anthropic a Supply Chain Risk: What Government Contractors Need to Know." https://www.mayerbrown.com/en/insights/publications/2026/03/pentagon-designates-anthropic-a-supply-chain-risk-what-government-contractors-need-to-know
CNN Business. (2026, May 1). "Pentagon strikes deals with 8 Big Tech companies after shunning Anthropic." https://www.cnn.com/2026/05/01/tech/pentagon-ai-anthropic
Bulletin of the Atomic Scientists. (2026, July). "Chinese nuclear weapons, 2026." https://thebulletin.org/premium/2026-07/chinese-nuclear-weapons-2026/
Ministry of Commerce of the People's Republic of China (MOFCOM). Export control notices (gallium and germanium, graphite, antimony, rare-earth-related items, etc.). https://www.mofcom.gov.cn/

If you are interested in improving export control operations or streamlining classification, review the feature overview in the TRAFEED product catalog (PDF), or contact us.