Ukraine tested AI-powered drones that killed without human control in battlefield milestone

Ukraine's one-time test used fully autonomous drones to kill Russian soldiers

Fully autonomous drones killed Russian soldiers during a battlefield test two years ago, according to a Ukrainian drone manufacturer. If true, the incident would represent another milestone in a war that has spurred unprecedented developments in military drones, robots, and AI-guided weaponry. The one-time test was revealed by Alexander Kokhanovskyy, CEO of the Ukrainian drone maker Aero Center, during an interview with New Scientist at a press event hosted by the Ukrainian embassy in London. Kokhanovskyy described the test using quadcopter drones that were preprogrammed to fly to a front-line area before activating an AI-powered "Terminator mode" that would seek out and attack any target in the given area. There was apparently no video feed or anything else to show what the "Terminator" drones targeted and attacked. But Kokhanovskyy told New Scientist that human-piloted drones sent to check out the aftermath found "a couple" of dead Russian soldiers, which led to the conclusion that the fully autonomous drones had killed them. A Ukrainian military commander told New Scientist that his drone pilots only use semi-autonomous systems that always have humans making crucial control decisions. He also described Ukraine's commitment to "international humanitarian law" while emphasizing that the military always exercises "great care in decision-making in order to prevent civilian casualties." The one-time nature of this experiment makes sense when considering the practical limitations of this approach, along with considerations regarding international humanitarian law. Sending fully autonomous drones to attack anything and everything in a given area without any human operator intervention requires careful preplanning and carries the risk of so-called "friendly fire" incidents or attacks on civilian noncombatants. It is also unclear how effective these fully autonomous quadcopter drones were in selecting and attacking targets compared to human drone pilots. There is currently no commonly agreed definition of what constitutes a lethal autonomous weapon system, according to the United Nations Office for Disarmament Affairs. But common characterizations describe weapon systems with the autonomy to "perform their functions in the absence of direction or input from a human actor." US Department of Defense policy has defined lethal autonomous weapons as "weapon system[s] that, once activated, can select and engage targets without further intervention by a human operator." The role of autonomous AI in drone arsenals Fully autonomous weapons with the capability to "accomplish goals independently or with minimal supervision in complex and unpredictable environments" are not yet a battlefield reality in the war in Ukraine, according to Kateryna Bondar, a former advisor to the government of Ukraine, in her report for the Center for Strategic and International Studies (CSIS) think tank in Washington, DC. But she highlighted a growing number of drones integrating certain autonomous capabilities for navigation and sometimes targeting, even if human operators maintain overall control. Ukraine and Russia are fielding many FPV drones for scouting and striking vehicles and even individual soldiers. These are typically controlled by trained drone pilots who wear virtual-reality goggles to see from the drone's point of view while aiming for enemy targets. There are also larger quadcopter or multirotor "bomber" drones that can carry heavier payloads for either supply runs or for dropping explosives on enemy targets at the front lines. Longer-range strike drones that resemble fixed-wing aircraft may incorporate more autonomous decision-making capabilities. In 2025, Russia launched hundreds of drones to attack Ukrainian cities each night, including Shahed drones originally provided by Iran and increasingly manufactured in Russia. Shahed drones are typically preprogrammed to fly automatically toward their targets with little autonomous decision-making capability. But some Shahed drone variants, such as the Geran-2, are equipped with smuggled Nvidia Jetson Orin microcomputers that provide onboard video processing and autonomous decision-making capabilities, including autonomous target recognition and retargeting. To defend against those attacks by Russian Shahed drones, Ukraine has deployed low- and high-tech air defense systems that include homegrown, inexpensive interceptor drones. Some interceptor drone systems are designed to autonomously fly to the intercept point and lock onto targets -- although a human operator is still required to perform initial target selection and initiate strike commands while always retaining the ability to cancel attacks, according to the Ukrainian government platform United 24. Meanwhile, Ukraine has gained the technological and manufacturing capability to launch more than 5,000 drone strikes against Russian targets at ranges exceeding 20 kilometers every month, according to the Ukrainian Ministry of Defense. These mid- and long-range strike drones necessarily rely heavily on autonomous navigation capabilities because of Russian electronic warfare systems that can interfere with human operator communication links, along with GPS jamming that can throw off GPS-guided weapons. Such AI-driven navigation has boosted the success rate of Ukrainian drone strikes from around 10 to 20 percent to 70 to 80 percent, Bondar wrote in her CSIS report. Overall, Ukraine's defense industry has focused on training small AI models on small datasets to run on the limited computing power of small and inexpensive chips, Bondar wrote. That approach has created AI-driven software for crucial autonomous functions -- including navigation and target recognition -- that can be packaged with standalone hardware modules for installation on small first-person view (FPV) drones, long-range strike drones, or even the gun turrets of uncrewed ground robots. So even if fully autonomous systems remain rare, expect more drones and robots to gain specific autonomous AI capabilities that complement human decision-making on the battlefield.

Ukraine used 10 AI-controlled 'Terminator' drones to kill Russian soldiers two years ago, marking first autonomous killings of humans -- autonomous killer quadcopters left 'everything dead' says senior Ukrainian defense industry figure

'We just launch it and we know everything will be dead,' said a drone maker and technology supplier. A watershed moment occurred on the battlefields of Ukraine in 2024 that we are only just hearing about today: the use of fully autonomous drones to kill humans. Ten fully autonomous quadcopter drones were sent to the front line by Ukraine with their AI-controlled 'Terminator Mode' engaged. "We just launch it and we know everything will be dead," a drone maker, Alexander Kokhanovskyy, told New Scientist at a press event hosted by the Ukrainian embassy. Human-piloted drones were later sent to recce the target area, and the machine-slain victims were concluded to include "a couple of soldiers, one truck." The New Scientist interview provides the most compelling evidence yet that humans have been killed at the sole discretion of an AI. This wasn't an accident of any kind, or an AI gone rogue; the 10 drones were sent to the front line purposely switched to an autonomous mode to search for and intercept targets. Significantly, "There is no connection to the drone at all, you cannot see the video, nothing... Everything it sees will be killed," explained Kokhanovskyy. So, human decision-making and judgment were removed for what is being characterized in the source report as a one-off test mission. That we are only hearing about it two years after the fact perhaps reflects on the gravity of engaging the Terminator Mode of the drones for (perhaps) the first time in an actual battle. A Rubicon moment for modern warfare? As the source notes, there is no official ban on using autonomous weapons that can kill without human intervention. However, last year the United Nations Secretary-General, António Guterres, called for this red line to be drawn. The argument is that removing human judgment from warfare risks human rights. Moreover, at this stage of the development of this technology, AI-driven systems are frequently in the news for the mistakes they make. Interestingly, Ukraine actually has a ban on fully autonomous final-stage targeting of humans by its drones. Other officials at the press conference where Kokhanovskyy spoke indicated that government decision-makers were in talks with defense companies about flexing the rules. However, it is likely not the first/only country to have crossed the Rubicon and allowed AI drones to inflict casualties on enemy soldiers. Academics speaking to the New Scientist offered some interesting angles on the news of Terminator Mode-engaged drones being let loose on the battlefield. One described the news as horrendous and a theft of human dignity. Another concluded that the 2024 AI-drone attack near the cities of Bakhmut and Chasiv Yar must have demonstrated that it is better to keep humans in the loop for military effectiveness. Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

How Ukraine Uses A.I. to Knock Deadly Russian Drones Out of the Skies

At a launch site deep in a pine forest in central Ukraine, five men in balaclavas removed a camouflage cover from a lightweight replica of a Russian-made Shahed drone. "Pull!" one man shouted, and the winged, triangular drone flew up over the trees. Standing nearby was a miniature black rocket with four propellers. This was the P1-Sun Long, one of Ukraine's first interceptor drones powered by artificial intelligence. Trained to find and knock down Shahed-type aircraft, it, too, took off. The P1-Sun reflects both the Ukrainian military's embrace of A.I. and the rapid evolution of its defenses against Shaheds. Russia fires the drones -- essentially flying bombs -- into Ukrainian cities in relentless daily waves, destroying infrastructure, killing civilians and sowing terror. Earlier in the war, Ukraine relied mostly on heavy machine guns, electronic warfare and occasionally missiles to bring down Shaheds. Last August, the Ukrainian military began to widely deploy interceptor drones piloted by humans, without A.I. Now, a major Ukrainian drone manufacturer, SkyFall, which conducted the recent test in the forest, says its interceptors have made dozens of A.I.-assisted strikes on Shahed-type drones since November, among thousands of interceptions overall. The interceptor drones are among a range of A.I.-powered weapons that have been deployed in recent months after being trained on the immense troves of data that the war has produced. These weapons include unmanned ground vehicles equipped with machine guns that use A.I. to help them quickly identify small quadcopters known as F.P.V. drones, because of the first-person view they offer to remote pilots. A.I. is also used in so-called terminal guidance systems, in which a weapon locks onto a target in its final approach and completes its strike without further human intervention. Many A.I. systems under development can autonomously identify objects like enemy vehicles. Some, in a more sensitive potential application, can single out people, like enemy soldiers. SkyFall is among those quietly testing people-targeting capabilities in Ukraine. A.I. developers and Ukrainian leaders say they are aware of the risks as the war spawns a technological revolution. As an era of fully autonomous lethal weapons approaches, human rights groups say that reducing life-or-death decisions to an algorithmic calculation is a threat to humanity. Want to stay updated on what's happening in Ukraine? Sign up for Your Places: Global Update, and we'll send our latest coverage to your inbox. At the same time, Ukraine is eager to use any technology that can increase its interception rate against Shaheds, and that means turning to highly automated air defenses. As of now, A.I. does only part of the work. Humans remain involved in critical aspects. SkyFall's A.I. system has been trained on more than 10,000 videos of interceptions of Shaheds. According to Brave1, a Ukrainian government defense technology hub, dozens of companies are using the videos to train A.I. systems to recognize Shaheds. SkyFall also regularly conducts exercises using replica targets, but it could not possibly fly enough decoys to fully train the system. During the recent demonstration in a small forest clearing, the pilot of the SkyFall interceptor, holding a remote control, guided it into the sky. The A.I. system was first to detect the Shahed decoy, long before the pilot could have. It marked the target with a green square on the pilot's screen. The pilot flew the interceptor drone toward the decoy until its shape was clear on his screen. Then he gave the auto-targeting system an order to follow it, and he let go of the controls. Once the interceptor had steered itself close enough to hit the Shahed, the pilot gave a final order, pressing the button to go in for the kill. The goal of the A.I.-driven system is to significantly shorten the time needed to detect and track enemy drones. Recently, the Ukrainian defense minister, Mykhailo Fedorov, said that a company participating in the Brave1 program had created a technology that automated 95 percent of the interception process. This system cannot yet initiate a launch automatically. It takes over, Mr. Fedorov said, after an operator selects a target and authorizes engagement. From there, it independently guides the interceptor toward the Shahed, autonomously identifies it and homes in. The company is MaXon, a defense technology start-up founded early last year. With even more autonomy, interceptors could stand ready to launch automatically after radar detects an attack. SkyFall says it is testing such a system. The advances may eventually allow one pilot to oversee several missions, instead of just one at a time, as is the case now, helping Ukraine's outmanned military. In part because of its personnel shortages, Ukraine is producing a wide range of aerial, sea and land drones. The hundreds of companies producing the drones aim to make them as inexpensive and as automated as possible. According to the head of autonomy and computer vision for SkyFall, who asked not to be named for security reasons, his company alone can make 50,000 interceptors a month. A bigger challenge is training enough skilled pilots; they remain on duty around the clock. Ukraine and other countries require highly autonomous systems to be able to operate at scale. Such development is important to Ukraine not just for its own defense, but also for its hopes of becoming a major defense exporter. After the United States and Israel began their war against Iran this year, those two countries and Persian Gulf nations used hundreds of costly interceptor missiles to shoot down inexpensive Iranian-made Shaheds. Ukraine quickly offered its cheap interceptor drones, without A.I. systems, as an alternative. Several Ukrainian companies have introduced remote-control technologies that allow missions to be piloted from Kyiv. Such offerings have helped Ukraine's president, Volodymyr Zelensky, cut security deals with Gulf states. In return, Mr. Zelensky has hoped to receive U.S.-made Patriot air defense systems, which Ukraine needs to shoot down Russian ballistic missiles. The Ukrainian leader has acknowledged the perils of A.I.-powered weapons, saying, "It's only a matter of time -- not much time -- before drones are fighting drones, attacking critical infrastructure and targeting people all by themselves." Weapons are evolving faster than humanity's ability to protect itself, Mr. Zelensky said in a speech to the United Nations General Assembly in September. As A.I.-enhanced drones got cheaper and proliferated, even small terrorist groups and cartels would use them in attacks, he said. Officials at companies like SkyFall and military commanders continue to stress the need for human confirmation before completing a kill. But at this point it is clearly an ethical, rather than a technological, limitation. The forest tests also involved applying A.I. to help F.P.V. drones find targets -- both equipment and personnel -- on the ground. As a dark green minivan drove in and out of the trees, a pilot designated points near it as a target. The A.I. automatically adjusted its aim onto the vehicle. In addition, the exercise involved identifying potential human targets: Members of the team walked around the clearing pretending to be Russians. Developers from SkyFall, eating sandwiches and cookies, watched on a monitor while a drone's A.I.-assisted targeting system located a colleague whose call sign was Forest. The system locked in on him and waited for the pilot's order to strike or stand down.