KHARKIV REGION, Ukraine—He goes by the military call sign “Mathematician.”
Before the war, he specialized in functional and complex analysis.
“I was like a freaky guy with long hair that lives in the forest and writes some magical stuff from differential equations,” he said, laughing, a hearty sound in a cold warehouse housing some of Ukraine’s most innovative drones.
He has the formula for the Fourier transform as a tattoo on his arm, and under that, something even more complex to mull over: the words “No Human Is Immortal,” a favorite quote of a close friend who died on the first day of Russia’s full-scale invasion in February 2022.
Nowadays he does very little math. Like many of his peers, he was suddenly and unexpectedly thrust into a military role after the Russians tried to take over his country.
But only a small slice of his time as an officer is spent on combat operations—mostly, he advises several units on drone technology. It’s a job that is half-R&D and half-fieldwork testing—a vital role in an area that is shaping the future of warfare.
Over a few short years, small drones have morphed from playthings—meant for wedding photography and weekend hobbyists—to deadly weapons of war, dozens of which can be coordinated for maximum effect. In fact, drones have become the defining weapon of the war in Ukraine, allowing a country with a meager air force, few missiles, and a scant navy to hit targets far beyond the frontlines.
Mathematician’s unit, part of the Khartiia Brigade station in the northeast corner of Ukraine, is engaged in some of the most cutting-edge drone innovation in the world. In December, the brigade pulled off the first ever air- and ground-drone assault on a Russian position. To pull it off, they had to come up with new methods of mission planning, tactics, and coordination procedures from scratch.
“It’s not very mature, it’s like a basic primitive [process],” said Mathematician, who declined to give his real name, a practice nearly universal among troops actively engaged in fighting because of security concerns.
Since the full-scale invasion began in 2022, the war has been a lopsided fight: Russia is much larger in both its economy and available manpower reserves. Drone technologies have given Ukraine one way to leverage its highly educated population.
Unmanned aerial drones fly hundreds of miles behind enemy lines to target Russian energy infrastructure; interceptors take down incoming kamikaze drones aimed at Ukrainian cities; and naval drones have been a key part of why one-third of the Russian navy’s Black Sea fleet has been disabled.
“We’re trying to scale it up,” Mathematician said, of their ongoing efforts so as to save Ukrainian lives. With drones, he points out, “we do not need to send [soldiers] close to the front line.”
‘Cheap tech’ and near-peer, decentralized innovation.
The incredible pace of innovation around drone tech signals that future wars between near-peer adversaries may be as much about technological innovation as they are about physical destruction.
At the beginning of the war, Ukrainians took whatever drone technology they could find off the shelf to eke out any advantage. One teenager became known as “drone boy” for flying a commercial drone to spy on Russian troops as they surrounded Kyiv—then passing on the intel to Ukrainian forces. Three years in, Ukrainians have gone from tinkering with Chinese-made commercial drones to developing brand new capabilities and tactics.
There are now drones that mine (and demine), and drones that accelerate in the air to intercept and kill incoming drones in a fit of drone-on-drone violence. Both sides have started mounting various anti-drone tools on the air frames: shotguns, for example, and even nets to entangle an adversary’s machines.
Ukrainians and Russians have developed dummy drones to trick enemy air defenses. And they’ve invented portable detection devices that sound when a drone is incoming, so as to warn the soldiers nearby .
New military technology in Ukraine has been driven by three key factors.
First, there is no motivation quite like knowing that people you care about can and will die if you don’t invent new solutions to battlefield problems. Innovators in Ukraine are constrained by a lack of resources and a lack of time—but that pressure has yielded the formation of hundreds of new domestic defense-tech startups that didn’t exist before the full-scale invasion.
Second, three years of war has made Ukraine the world’s leading laboratory for real-world military weapons testing. The chance to engage in rapid battlefield iteration has meant that tinkering and improvement is not just theoretical but practical and combat tested.
The third aspect explaining the speed of innovation is the decentralized way that the Ukrainian military has worked since the war began. While there is still a centralized defense procurement system, individual soldiers and units are also involved in finding tech and equipment to suit local needs.
Often, it comes down to who you know. Soldiers on the frontline without protection from Russian drone attacks might tell their parents, who might contact an NGO, who engages with inventors to build an anti-drone system. Ukraine is unique in that charities are involved in military procurement, purchasing, and shipping lethal and non-lethal military aid to the frontlines.
Requests for help also come from the unit level, where different military units will reach out to people they know, with specific requests suitable for the particular environment in which they’re fighting. These decentralized networks drive customization, creativity, and new inventions.
The resulting innovation—taking place in a period of scarce financial resources and time—has resulted in “cheap tech”—low-cost, good-enough machines that aren’t of the standard of American designs, but still get the job done.
Electronic warfare and the cycles of counterinnovation.
The success of drones has naturally led to efforts to reduce their effectiveness on the frontlines. Electronic warfare—usually in the form of using powerful signals to disrupt the connection between pilot and drone—has become ever-present on the battlefield.
Russia has a special relationship to EW, developed during the Cold War space race. It even has a national day of electronic warfare each year, to mark its status as the first country ever to use the concept in combat (in the early 20th century, during the Russo-Japanese war). It turns out that the same types of technology needed to block BBC broadcasts into the Soviet Union are the types of technologies useful for disrupting battlefield drone communications.
“The rapid evolution of drone warfare has outpaced our thinking about how to incorporate it into our existing laws of warfare, or our frameworks of ethics. Ukraine, fighting for its very existence, hasn’t the time to consider it—engineers build, or people die; that’s the bottom line. Its foreign partners do have that luxury.”
Each new development leads to new efforts to counter them: Both sides are now trying to innovate around the problems caused by electronic warfare by adopting yet other new technologies.
In some cases, physical fiber optic cables, running up to 20 kilometers long, tether flying drones to their pilots. With the signals running through the lines, rather than through the air, electronic warfare is ineffective.
This explosion in new ideas is reflected by a surge in the number of drone companies and drone orders. According to the Ukrainian government, 120 drone manufacturers and some 500 drone organizations now dot the Ukrainian economic landscape, producing 1 million drones annually—many of which are not publicly registered.
There’s also a real effort to increase automation in drones. Increasing independent decision-making done by the drone itself reduces its susceptibility to electronic warfare disruption: Drones that are programmed to navigate, deploy weapons, or carry out other operations on their own are more immune from attempts to block communications between drone and pilot.
Automation creates enormous potential for new offensive uses for drones which can spare human lives. But they also open up a whole new category of ethical and legal questions.
The first all-drone assault.
The Khartiia Brigade’s logo is meant to be a representation of a bow and arrow, but if you squint, it looks like a swarm of drones.
In the early morning hours, on the first week of December 2024, dozens of ground and aerial drones belonging to the unit took off from a forward position in the northeast of Ukraine, near the Russia-Ukraine border.
The fleet contained ground kamikaze drones, ground turret-mounted drones, aerial surveillance drones—even a flying drone with a rifle mounted on it. No humans physically accompanied them on this all-drone assault.
Around 100 soldiers took part in the planning, preparation, and execution of the mission. Pilots, physically located in different bases, coordinated the attack. As they watched on their screens, the drones sent back a live broadcast of what was happening in the field, which the pilots used to coordinate their actions.
Frankly, they were expecting the attack to be a failure. It was the first time that the Ukrainians had launched an assault using both aerial and ground drones in tandem, and they expected there to be some massive kinks to work out.
“It was like a testing mission to understand how we need to modify technical side, organization side of this whole area,” Mathematician recalled.
The higher-ups anticipated losing up to half of the drones midway through the mission. They knew they weren’t ready to try an assault like this on a massive scale but were trying to understand how to do it for the first time.
Against all their expectations, the attack turned out to be a success. The drone operators, listening in on intercepted Russian communication, could hear the chaos and confusion that ensued due to the unprecedented nature of the attack. The Russian forces abandoned their entrenched position, and Ukrainian troops later came in to capture it.
The Khartiia Brigade sees this whole operation as a test on whether larger drone assaults can replace infantry-led “meat grinder” operations, in which large numbers of soldiers are killed and wounded to capture enemy-held areas.
“I still want to create some system where we have some pilots that are remote, in some bunker with big monitors, with coffee and a sofa, [so that there can be] five guys that can operate some 20 [ground drones],” said Mathematician.
And over time, as autonomy increases, and a single pilot can operate multiple drones seamlessly, the unit will try to scale up the number of drones used in the attacks, increase the mobility of the ground drones, and improve the connections between pilots and drones.
“We can do this with a few pilots, and do a lot of missions [where] we do not need to send [people] close to the front lines,” the military officer said.
The ethics of lethal drone autonomy.
Isaac Asimov, in his science fiction universe that imagined sentient robots, set out a first law that every robot must follow: “A robot may not injure a human being or, through inaction, allow a human being to come to harm.”
Many of his stories revolved around how robots would find loopholes in the laws of robotics, leading to harm or death. On a trip shadowing a drone unit in the Donbas, a pilot controlling a deadly flying machine quipped to me, “We have already violated the first rule: Never kill.”
Ukraine is now a laboratory of drone inventions, almost all of it, in some way or another, aimed at increasing lethality. Humanity has skipped trying to program non-killing instincts into robots and gone straight to developing contraptions that kill as if on instinct. We may soon end up in a world that is even worse than what science fiction could conjure up.
The rapid evolution of drone warfare has outpaced our thinking about how to incorporate it into our existing laws of warfare, or our frameworks of ethics. Ukraine, fighting for its very existence, hasn’t the time to consider it—engineers build, or people die; that’s the bottom line.
Its foreign partners do have that luxury. We from outside Ukraine have the uncomfortable responsibility of contemplating the ethical considerations of how to use drones, especially as drone autonomy is getting more advanced by the day.
Consider the following scenario: A drone is sent north to a “kill box” of 10 square miles. It has been taught through machine learning—trained by thousands of photos—how to identify Russian tanks. If it observes a Russian tank in that area, it is programmed to approach it and explode.
This solves the practical question of how to avoid electronic warfare. By acting relatively independently, no signal needs to pass from pilot to drone. So none can be intercepted or disrupted.
But this now poses new ethical questions. Let’s say that as the drone approaches a Russian tank, a nearby Russian soldier decides to surrender, knowing that it has no tools to fight back against the drone. The drone has no programming to react to this input.
It explodes, striking the tank and killing the soldier. Is this a war crime? Can a soldier surrender to a drone? And what if the drone were not properly programmed to assess collateral damage, and kills children standing at a hospital nearby?
Furthermore, the responsibility for the act has been dispersed: If a war crime has been committed, who is responsible—the pilot, the pilot’s commander, the manufacturer, the programmer?
“I think the end user will be to blame. That is, the person who gave it these commands, or taught it this command. … if you teach him badly, he will do badly,” said “Happy,” a colleague of Mathematician in the Khartiia Brigade.
While it is understandable that ethics take a back seat to innovation given the urgency of the war for Ukrainians, many Western defense firms are also trying to get into the action: Drone companies from all over are exporting their tech to Ukraine to slap the label “combat tested” on their products.
But in the absence of any similar existential threats, foreign governments and companies need to lead the effort to consider an ethical approach to lethal drone autonomy, the kind of thinking that the Ukrainians can’t afford to do in this time of their emergency.
What’s next for drone innovation.
Most governments outside of Ukraine have done little thinking about how to integrate these tools into average military units to reduce the risk to human soldiers. A future where war is conducted mostly by machines, rather than human troops, is now within sight. And in the interim period, militaries should consider how to mitigate potential adversaries’ drones while building up their own stock.
Meanwhile, there is as much potential for drones to be instruments of peace—think automated demining drones—as there is for them to be instruments of violence.
The same drones that carry munitions can carry fresh blood supplies to far-off, austere locations for transfusions. Ground drones that carry mounted machine guns could just as easily carry injured casualties from a burning building. Hovering aerial drones could fly to ensure internet communications after a natural disaster, rather than providing surveillance of the battlefield.
The development of drones also represents tremendous economic hope for war torn Ukraine in the future, once the war ends. The human capital that has grown rapidly there in just three years of conflict can be tapped into and exported for the betterment of the democratic world—whether for civilian or military uses.
But whether drones are used for good or violence depends on us, the operators.