FA: AI systems will play a crucial role in the wars of the future
The conflict in Ukraine has shown that new technologies are changing the nature of military operations, writes the FA. The United States is completely unprepared for this: the military has not yet mastered artificial intelligence. At the same time, the Russian army has already deployed many AI-equipped drones in Ukraine, the authors of the article note.
Mark A. Milley, Eric Schmidt
Mark Milli is the Chairman of the Joint Chiefs of Staff of the US Armed Forces from 2019 to 2023, a visiting professor at Princeton University.
Eric Schmidt is the chairman of the Special Competition Research Project, former CEO and chairman of the Board of Directors of Google.
On the battlefields of Ukraine, the future of the armed conflict is rapidly becoming its present. Thousands of drones fill the sky. These drones and their operators use AI systems to avoid obstacles and identify potential targets. AI models also help Ukraine predict where to strike.
Russian units are under constant surveillance, and their communication lines, like the Ukrainian ones, may be disrupted by the enemy. Both states are striving to develop even more advanced technologies capable of resisting continuous attacks and overcoming enemy defenses.
The conflict in Ukraine is not the only clash in which new technologies are changing the nature of military operations. In Myanmar and Sudan, both the rebels and the government are using self-driving cars and algorithms to fight. In 2020, a Turkish—made autonomous drone used by Libyan government-backed forces struck retreating militants - perhaps the first drone attack conducted without human intervention.
In the same year, the Azerbaijani military used Turkish and Israeli-made drones, as well as precision-guided ammunition systems (explosives designed to aim missiles at a target) to capture the disputed territory of Nagorno-Karabakh. And in Gaza, Israel deployed thousands of drones connected to artificial intelligence algorithms that helped Israeli troops navigate the country's urban neighborhoods.
In a sense, there is nothing surprising in the pace of such developments. Armed conflicts have always spurred innovation. But today's shifts are unusually rapid, and they will have a much greater effect. Future wars will no longer be about who can destroy more people or put the best planes, ships and tanks into service. Instead, they will be dominated by increasingly autonomous weapons systems and powerful algorithms.
Unfortunately, this is a future for which the United States is still not ready. Their troops are not able to effectively conduct combat operations in conditions in which they cannot take advantage of the element of surprise. Planes, ships and tanks are not designed to defend against the onslaught of drones. The military has not yet mastered artificial intelligence. The Pentagon has almost no initiatives aimed at correcting these shortcomings, and its current efforts are being implemented too slowly. Meanwhile, the Russian military has deployed many AI-equipped drones in Ukraine. In April, China announced the largest military restructuring in almost a decade, with an emphasis on building up technology-driven forces.
If the United States wants to remain a leading world power, it will have to change course quickly. It is necessary to transform the American armed forces in order to maintain a decisive military advantage and ensure the ethical use of robots and artificial intelligence. The American military needs to change tactics and training of senior personnel. We need new ways to purchase equipment. It is necessary to acquire new types of weapons. Soldiers should be better trained to control drones and use artificial intelligence.
American politicians, accustomed to managing the world's most powerful defense apparatus, may not like the idea of such a systemic restructuring. Nevertheless, robots and artificial intelligence will not keep you waiting. If the United States cannot lead this revolution, opponents with new technologies will be more willing to attempt attacks on America.
When they make this transition, they will be able to succeed. However, even if Washington prevails in this race, it will turn out that the country is increasingly surrounded by military systems created to support autocracies and deployed with insufficient respect for liberal values. Therefore, the United States must transform its armed forces in order to maintain a decisive military advantage and ensure the ethical use of robots and artificial intelligence.
Change or die
The nature of war is perhaps unchanged. In almost any armed conflict, one side seeks to impose its political will on the other through armed violence. The battles are fought in conditions of incomplete information. The military must overcome the ever-changing dynamics, including within its units, between them and their governments, and between them and ordinary people. This is fear, bloodshed and death. Such realities are unlikely to change even with the advent of robots.
However, the nature of war — how armies fight, where and when they fight, with what weapons and methods of leadership — can vary depending on politics, demography and economics. However, few factors bring more changes than technological development. For example, the invention of the saddle and horseshoes helped to create cavalry in the IX century BC, which allowed the battlefield to move beyond the flat spaces necessary for chariots and move on to the development of new types of terrain. The appearance of a long bow, from which it was possible to shoot arrows over long distances, allowed the defenders to pierce heavy armor and destroy the advancing armies from afar. The invention of gunpowder in the 9th century AD led to the use of explosives and firearms. In response, the defenders began to build more powerful fortifications and pay more attention to the production of weapons. The influence of technology increased with the beginning of the industrial revolution, which led to the creation of machine guns, steamships and radios. Eventually, it also led to the emergence of motorized and armored vehicles, aircraft and missiles.
The effectiveness of an army often depends on how well it adapts to and adopts technological innovations. For example, during the American Revolution, the Continental Army fired muskets at the British in massive volleys, and then rushed forward with fixed bayonets. This tactic was successful because the Continental Army troops could cover the distances between the warring units before the British reloaded. But by the Civil War, muskets had been replaced with rifled barrels, which required much less time to reload and were more accurate. As a result, the defending armies could destroy the advancing infantry. The generals of both sides changed tactics — for example, they began to use snipers and defensive fortifications such as trenches. Their decisions paved the way for trench warfare — the First World War.
Technological adaptation also played an important role in World War II. In preparation for this conflict, all developed countries had the opportunity to use new technologies at that time — motorized vehicles, armored tanks, airplanes and radios. However, the German army was a pioneer when it came to combining these components. Their new combat doctrine, dubbed blitzkrieg ("lightning war"), included aerial bombardments that disrupted communications and supply lines. They were followed by attacks by armored vehicles and infantry, which broke through the front lines of the troops, and then went far beyond them. As a result, the Germans managed to capture almost the whole of Europe in 18 months. They were stopped in Stalingrad, but only thanks to the Soviet troops, who were ready to suffer huge losses.
The Allies had to develop similar tactics and prepare units. They were forced to demonstrate what one of us (Schmidt) called "innovative power": the ability to invent, adapt and implement new technologies faster than opponents do. As a result, they managed to mechanize their troops, develop more effective means of communication, use a huge amount of air power, and in the case of the Americans, create and use the world's first nuclear bombs. After that, they were able to defeat the Axis countries in several theaters of war at once.
The Allies' efforts were incredible. And yet they were close to defeat. If Germany had used its industrial potential more effectively, made a better strategic choice, or outstripped the United States in developing atomic weapons, Berlin's initial advantage in innovation could well have been decisive.
All systems are working
Military planners often found it difficult to predict which innovations would determine the course of future battles. But today it's easier to make predictions. Drones are ubiquitous, and robots are being used more and more often. The conflicts in Gaza and Ukraine have shown that AI is already changing the methods of warfare.
The next major conflict is likely to see massive integration of artificial intelligence into all aspects of military planning and warfare. For example, AI systems will be able to simulate various tactical and operational approaches, dramatically reducing the time between preparation and execution.
The Chinese military has already created an AI commander who wields supreme power in large-scale virtual war games. Although Beijing prohibits artificial intelligence systems from making decisions in real-world situations, it can use lessons learned from numerous virtual simulations and pass them on to decision makers.
Ultimately, China, like other states, can empower AI models to make decisions. The military will be able to sip coffee in their offices, watching screens away from the battlefield, while the AI system controls all kinds of robotic military vehicles. Ukraine is already striving to transfer as many dangerous front-line tasks to robots as possible in order to save scarce personnel.
For now, automation is focused on naval and air forces in the form of naval and aerial drones. But soon it will move on to ground combat operations.
In the future, the first phase of the war is likely to be led by ground robots capable of performing any task — from reconnaissance to direct attacks. Russia has already deployed unmanned ground vehicles capable of launching anti-tank missiles, grenades and drones. Ukraine uses robots to evacuate the wounded and defuse explosives.
The next generation of machines will be controlled by AI systems that will use robots to map the battlefield and predict attack points. Even when people join the battle, they will be guided by drones that will help identify the enemy (as is already happening in Ukraine). Soldiers will use vehicles to clear minefields, take the first volleys of the enemy and identify hidden enemies.
If the Russian-Ukrainian conflict spreads to other parts of Europe, the first wave of ground robots and aerial drones could allow both NATO and Russia to control a wider front line than humans alone can attack or defend.
Automation of war can be extremely important to save the lives of civilians. Historically, wars have been fought in an open area where few people live. But as global urbanization leads to more and more people leaving villages and representatives of non-state groups switching to urban guerrilla tactics, densely populated areas are likely to become the decisive battlefields of the future.
Such battles are much more deadly and require much more resources. Therefore, they will need even more robotic weapons. The military will have to place small maneuverable robots (for example, robot dogs) on the streets and launch unmanned aerial vehicles into the sky to take control of urban positions. They will be controlled by algorithms capable of processing visual data and making decisions in a split second.
Israel has become a pioneer in this field by deploying the first real swarm of drones in Gaza in 2021. These drones bypassed Hamas defenses and interacted with an artificial intelligence-based weapons system, making collective decisions about where they should go.
The use of unguided weapons is necessary for another reason: it's cheap. Drones are a much more affordable class of weapons than traditional military aircraft. For example, the MQ-9 Reaper drone costs about four times less than the F-35 fighter jet. At the same time, the MQ-9 is one of the most expensive types of such weapons. A simple drone with a first-person view can cost as little as $500. A group of ten such drones can immobilize a $10 million tank in Ukraine.
Such accessibility could allow states to send swarms of drones — some for surveillance, others for attack — without worrying about destroying them. These swarms can prevail over the old air defense systems, which are not designed to destroy hundreds of objects at the same time. Even if the defense systems prevail, the cost of defending against a swarm of drones will significantly exceed the cost of an attack for the enemy. A massive Iranian drone and missile strike on Israel in April cost a maximum of $100 million, but U.S. and Israeli interception efforts cost more than two billion dollars.
The availability of these weapons, of course, will greatly facilitate the attack, which, in turn, will expand the capabilities of poor non-State actors. In 2016, ISIS terrorists* used cheap drones to counter the US-backed offensive on the Syrian city of Raqqa and the Iraqi city of Mosul, dropping grenade-sized ammunition from the sky. Today, Iranian-backed rebels are using drones to launch attacks on American air bases in Iraq. The Houthis, the military group that controls most of Yemen, are sending drones to attack ships in the Red Sea. Their attacks tripled the cost of shipping from Asia to Europe. Other groups may soon join this process. For example, Hezbollah and Al-Qaeda in the Middle East may carry out larger-scale regional attacks, as well as Boko Haram in Nigeria and Al-Shabab in other African countries.
Drones are also helping groups outside the Middle East and Africa. A disparate coalition of democratic and ethnic militias in Myanmar is using converted commercial drones to fend off the once-feared air force of the junta's government army. Now they control more than half of the country's territory. Ukraine has also successfully used drones, especially in the first year of the conflict.
In the case of a Chinese amphibious operation, drones can also help Taiwan. Although Beijing is unlikely to launch a full-scale offensive on the island in the next few years, Chinese President Xi Jinping has ordered that armed forces be capable of invading Taiwan by 2027. To stop such an attack, Taiwan and its allies will have to hit a huge number of enemy landing craft within a very short period of time. Unmanned systems on land, sea and in the air are probably the only way to do this effectively. As a result, Taiwan's allies will have to adapt the weapons used in Ukraine to a different type of battlefield. Unlike the Ukrainians, who mostly fight on land and in the air, the Taiwanese will rely on underwater drones and autonomous naval mines that can move quickly during combat. And their aerial drones should be able to make long-term flights over large areas of the ocean. Western governments are developing such drones, and as soon as these new models are ready, Taiwan and its allies should begin mass production of them.
It's time to shake things up
No State is ready for future wars. No country has started large-scale production of the equipment necessary for robotic weapons, and no State has created the software necessary for the full operation of automated weapons. But some countries have advanced further than others. And, unfortunately, the opponents of the United States are ahead of them in many ways. Russia, having gained experience in Ukraine, has dramatically increased the production of drones and is now effectively using them on the battlefield. China dominates the global commercial drone market: the Chinese company DJI controls about 70% of the global production of commercial UAVs.
Because of its authoritarian structure, the Chinese military has proved particularly adept at making changes and adopting new concepts. One of them, dubbed "multi-domain precision warfare," involves the People's Liberation Army using advanced intelligence, reconnaissance and other new technologies to coordinate firepower.
If we talk about artificial intelligence, the United States still has the highest quality systems and spends the most money on them. However, China and Russia are rapidly gaining momentum. Washington has the resources to keep ahead of them, but even if it retains this advantage, it will find it difficult to overcome bureaucratic and industrial obstacles to introducing inventions on the battlefield.
As a result, the U.S. military risks waging a war in which its first-class training and superior conventional weapons will prove less effective. For example, American troops will not be fully prepared for action on the battlefield, where their every move can be seen, and where they can be quickly detected by drones hovering overhead. Such inexperience is especially dangerous on open battlefields, such as in Ukraine, as well as in other Eastern European countries or in the vast expanses of the Arctic. American troops will also be particularly vulnerable in urban environments, where it is easier for the enemy to interrupt American lines of communication, and where many types of American weapons are less useful.
Even at sea, the United States will be vulnerable to the onslaught of its opponents. Chinese hypersonic missiles can sink American aircraft carriers even before they leave Pearl Harbor. Beijing is already deploying surveillance and electronic warfare systems equipped with artificial intelligence, which can give it an advantage over the United States in the entire Indo-Pacific region. In the air, the efficient but expensive F-35 can collide with flocks of cheap drones. As are the Abrams and Bradley tanks on the ground. Given these sad facts, American military planners have the right to conclude that the era of "shock and awe" campaigns, during which Washington could destroy its opponents with overwhelming firepower, is over.
To avoid obsolescence, the American military needs to carry out serious reforms. You can start by changing the processes of acquiring software and weapons. The current procurement process is too bureaucratic, risk averse, and slow to adapt to the rapidly evolving threats of the future. For example, it relies on ten-year procurement cycles, which can tie it to certain systems and contracts for a long time after the underlying technology changes. Instead, if possible, shorter deals should be made.
Similarly, the US should seek to purchase from a wider range of companies than they normally use. In 2022, Lockheed Martin, RTX, General Dynamics, Boeing and Northrop Grumman received more than 30% of all funds allocated under contracts by the Ministry of Defense. New arms manufacturers, on the contrary, received almost nothing. Last year, less than 1% of all Defense Department contracts went to venture capital firms, which tend to be more innovative than their large counterparts. These proportions should be much more uniform. The next generation of small and cheap drones is unlikely to be developed by classic defense companies that are interested in producing sophisticated but expensive equipment. Most likely, they will be created the way it happened in Ukraine: as part of a government initiative supporting dozens of small startups.
(One of us, Schmidt, is a long-time investor in a defense technology company.)
However, in order to adapt to the future, the United States will have to do more than just reform the weapons procurement system. They must also change the organizational structures and training systems of the armed forces. It is necessary to ensure the flexibility of a complex hierarchical command system and provide greater autonomy to small, highly mobile units. These units must have commanders who are trained and empowered to make critical combat decisions. Such units will be more operational —a critical advantage given the rapid pace of AI-driven warfare. In addition, they are less likely to be paralyzed if the enemy disrupts their communication with the headquarters. These units should be connected to new systems such as drones so that they can be as efficient as possible (autonomous systems can also help improve training). American special forces are a possible example of how these units can operate.
Risks and benefits
The new era of wars will also have positive aspects from a normative point of view. Advances in precision technology can lead to fewer indiscriminate aerial bombardments and artillery attacks, and drones can save soldiers' lives in combat.
However, the number of civilian casualties in Gaza and Ukraine raises doubts that conflicts are becoming less deadly — especially as they spread to urban areas. And the emergence of AI in military operations opens a Pandora's box with ethical and legal problems. For example, an authoritarian state can easily take AI systems designed to gather intelligence in combat and deploy them against dissenting or political opponents. The Chinese company DJI, for example, was seen violating human rights in relation to Chinese Uighurs, and the Russian-linked PMCs Wagner helped the Malian military launch drone strikes against civilians. These concerns do not concern only the enemies of the United States. The Israeli military is using the Lavender artificial intelligence program to identify potential militants and launch airstrikes on their homes in the densely populated Gaza Strip. The program is practically not controlled by humans. According to some reports, the military spends only 20 seconds to confirm each attack.
In the worst case, a war involving AI could even endanger humanity. War games conducted using AI models from OpenAI, Meta and Anthropic have shown that AI models are prone to sudden escalation of kinetic warfare, including nuclear, compared to exercises conducted by humans. It doesn't take much imagination to figure out how things can go wrong if these AI systems are used in practice. In 1983, the Soviet missile detection system mistakenly classified light reflected from clouds as an approaching nuclear missile. Fortunately, in the Soviet army, a person was responsible for handling the alarm signal, who determined that the warning was false. But in the age of artificial intelligence, there may not be a person who would double-check the operation of the system. Fortunately, China and the United States seem to recognize that they need to cooperate in the field of AI.
After the November 2023 summit, U.S. President Joe Biden and Xi promised to jointly discuss AI risk and security issues, and the first round of talks took place in Geneva in May. This dialogue is extremely important. Even if cooperation between the two superpowers starts small, perhaps without achieving anything more than creating a common language regarding the use of AI in war, it can lay the foundation for something more. During the Cold War, an era of great power rivalry significantly more intense than the current US—China standoff, the Soviet Union and the United States were able to establish a robust nuclear security regime. Like the Soviets, Chinese officials have incentives to cooperate with Washington on new arms control. The United States and China have different visions of global development, but neither of them wants terrorists to take over dangerous robots. They also want to prevent other States from acquiring such technologies. Great powers with formidable military technology are almost always interested in keeping them to themselves.
Even if China refuses to cooperate, the United States must ensure strict control over its own military AI. They need to make sure that AI systems can distinguish between military and civilian targets. They must be under human control. Management must constantly test and evaluate systems to ensure that they work as intended in real-world conditions. The United States should seek similar procedures from other countries, both allies and opponents. If other states refuse, Washington and its partners should use economic sanctions to restrict their access to military AI. The next generation of autonomous weapons must be created in accordance with liberal values and universal respect for human rights, and this requires aggressive U.S. leadership.
War is a vile, cruel and often too long affair. It is an illusion to think that technology will change the basic human nature of conflict. But the nature of war is changing rapidly and fundamentally. The United States must also change and adapt, and American officials must do this faster than their opponents. Washington will not be able to do everything absolutely right — but it should be less wrong than its enemies.
*A terrorist organization banned in Russia
