Artificial intelligence-driven autonomous lethal weapon systems may be transitioning from science fiction to reality. In the deadlock of the Russia-Ukraine battlefield, Kiev and Moscow are crossing ethical red lines in their quest for war dominance.
Simultaneously, the Pentagon has begun investing billions of dollars to promote autonomous combat transformation. In this nearly uncontrollable technological arms race, the “Human-in-the-Loop” principle, once considered a safety baseline by the international community, may be crossed, and all of this is quietly unfolding on the plains of eastern Ukraine.
“Autonomy” is the ability granted by AI. Ethical principles dictate that machines should not decide life and death for humans. Technology experts have warned repeatedly, “Once this Pandora’s box is opened, it will be hard to close.”
Yet, in actual combat, it is the Russia and Ukraine sides, battling fiercely on the Ukrainian battlefield, that are driving the accelerated development of unmanned combat systems. Will autonomous weapons see explosive development in 2026? The Russia-Ukraine battlefield provides the best window for observation.
On September 26, 2026, Mykhailo Fedorov, former Ukrainian Defense Minister and key proponent of unmanned combat systems, proposed the “Robot Legion” plan at the IT Arena technology conference in Lviv, Ukraine.
Fedorov believes that as Russia accelerates the development of autonomous drones and ground combat systems, Kiev cannot afford to be entangled in the ethical controversies of AI weapons for too long. He pointed out that AI robots could become a key factor in the next stage of war technology development.
He argued that once the first batch of robots is deployed, this trend will be nearly unstoppable and difficult to reverse. For Ukraine, the key lies in deploying ahead, gaining battlefield advantage, and then leading the future direction and rules of autonomous combat.
With Russia advancing unmanned tank and autonomous drone technologies simultaneously, the competition is no longer just about weapon performance but also about who can establish new combat modes first and who will dominate the future battlefield.
On October 14, 2026, US Defense Secretary Hagseth discussed plans to massively expand autonomous weapons and establish a new four-star Autonomous Warfare Command (AUTOWARCOM) at the annual AUSA Conference in Washington.
Hagseth emphasized that the US military must integrate autonomous systems, robots, superintelligence (SI), and space capabilities and closely collaborate between defense procurement personnel and frontline combat troops to continuously improve and innovate.
The goal of this policy is to shift autonomous combat from dispersed technical research and procurement work across military branches to coordinated construction. The Pentagon is establishing organizational and executive structures for the Autonomous Warfare Command through “Project Agincourt.”
The US Army is separately pushing for the establishment of the Army’s “Future and Autonomous Systems Command” (FASCOM), integrating autonomous technology development within the Army. These measures show that the US military is trying to shorten the distance between technology research, procurement, and battlefield application, accelerating the deployment of autonomous weapons and robot combat capabilities.
Currently, the 2027 fiscal year defense spending bill has not been approved by Congress, but the Pentagon has proposed a plan to invest nearly $55 billion in work related to the Defense Autonomous Work Group.
Hagseth emphasized that a complete budget request is crucial for driving internal reforms within the Department of Defense. He stated that this is a one-time investment that can affect an entire generation, its effects lasting for years and continuing to impact the entire joint forces for decades to come.
Mykhailo Fedorov envisions a future where AI-driven robots play a central role on the battlefield.
He noted that Ukraine is currently engaged in a war with Russia that determines the nation’s survival, with the survival crisis pushing the ethical and moral risks of AI to a secondary position. The later Ukraine deploys AI robots, the more time it gives Russia, which is also testing its own unmanned systems.
Fedorov is a new generation Ukrainian tech-savvy political figure and a key figure in promoting Ukraine’s digital reform and unmanned combat systems.
In 2019, he assisted Zelensky in winning the presidential election, then served as the Minister of Digital Transformation, promoting the e-government platform “Diia” and securing Western tech company support post the 2022 full-scale Russian invasion, facilitating services like Starlink and other digital technologies to meet Ukraine’s wartime needs, while advancing drone procurement, battlefield data integration, and defense technology innovation.
In January 2026, he assumed the position of Ukraine’s Defense Minister, attempting to further push military system reform, but was replaced by Zelensky after about 6 months in office. Even after leaving the government, Fedorov continued to be involved in defense technology and collaborated with companies like Palantir Technologies to promote autonomous combat robots and AI military applications.
The use of artificial intelligence in the defense sector has always been a controversial topic. Many officials in public and private sectors are concerned that introducing AI-driven military systems into real battlefields may raise moral and ethical issues. Considering the significant target identification errors during the war on terror, people are worried that autonomous weapon systems might mistakenly classify civilians or friendly forces as targets.
The most contentious issue of AI entering the military field lies in whether machines, not solely assisting humans in analyzing intelligence, identifying targets, and planning attacks, but starting to choose targets, determine the timing of attacks, and even continue carrying out strike missions after losing communication contact, how much control humans can still exercise over lethal force?
In the unique circumstances of the Ukraine battlefield, the priority of this issue is rapidly diminishing. For Ukraine, survival is the foremost requirement, and autonomous combat technology is primarily a means of survival.
The concept of the “Robot Legion” aims to develop robots and artificial intelligence from the lab to have tangible capabilities for reconnaissance, logistics, mine clearance, transportation, fire support, and other combat missions. This vision encompasses various robots, including humanoid robots capable of mimicking human actions, allowing these machines to undertake the most dangerous work, reducing frontline casualties. Ultimately, different autonomous platforms can operate collaboratively according to mission requirements, forming comprehensive combat capabilities.
Currently, wheeled or tracked unmanned ground vehicles are used by the Ukrainian military for transporting supplies, casualties, explosives, and carrying weapons for certain combat missions.
From September to October 2026, the Ukrainian military began integrating grenade launchers and rocket-propelled grenades onto unmanned ground vehicles for attacking Russian positions, allowing soldiers to deliver firepower from a relatively safe distance.
These advancements indicate that robots have begun extending from logistical and support platforms to direct combat capabilities. However, these capabilities are still under remote human control and cannot yet be classified as fully autonomous lethal weapons due to limitations in their usage and autonomy.
Fedorov hopes to witness the first casualties caused by humanoid robot soldiers within six months and to achieve ground assaults that do not require direct human involvement within twelve months.
From remote-controlled robots to autonomous killing systems, there is still a distance to cover. The autonomy of military robots can be divided into several levels. The most basic is unmanned systems controlled by humans, such as first-person view drones; then those with route planning, obstacle avoidance, and the ability to continue missions in case of communication disruption; following are systems that use AI to analyze sensor data, identify and track targets; lastly, systems that can autonomously attack targets, which also represent the most controversial part.
At this stage, Fedorov’s goal is to have robots undertake more combat missions, reduce Ukrainian military casualties, without completely removing humans from the chain of harm. However, if robots eventually achieve autonomous target selection and attack capability, the risks of war would change accordingly. Operating without human intervention may expedite decision-making, but not necessarily ensure decisions are correct. If systems swiftly turn faulty judgments into attack actions, it could lead to tragic consequences on the battlefield.
When one side can use machines to carry out more high-risk tasks while the other must endure sustained high personnel losses, decision-makers might prioritize how to reduce military casualties. This might result in technologies that originally required long-term ethical review being haphazardly deployed to the battlefield due to pressing wartime situations.
Fedorov’s view is that whoever gains the advantage in autonomous combat could potentially dictate the rules of future battlefields. He worries that if Russia establishes a mature autonomous weapon system first, Ukraine will be forced to fight in the operational environment set by the adversary. To avoid being passive on the battlefield, this could lead to an uncontrolled arms race in autonomous lethal weapons between both sides.
Currently, Fedorov’s ambitious “Robot Legion” plan remains in a stage requiring humans to transmit images via radio or satellite communication, identify targets, and then decide whether to attack. The issue is that communication systems connecting operators with drones are often subject to electronic interference or complete interruption. This shift towards unmanned systems capable of autonomous navigation, target tracking, identification, and completing attack missions independently when human operation commands are disrupted represents the future development direction.
On October 12, 2026, Fedorov showcased a new high-speed intercept drone. He mentioned that this drone is still operated by humans, but the development team plans to integrate artificial intelligence to enable autonomous discovery and interception of Russian jet attack drones without requiring operator intervention. This development reflects the practical need of the battlefield for autonomous technology. Faced with communication disruptions in electronic warfare environments, autonomous weapons that can conduct attacks independently without relying on communications may be the most direct solution.
However, the fact that unmanned systems continue executing missions in the absence of human control does not mean they should have the authority to independently decide to attack humans. This is the key issue that sparks major ethical and legal controversies surrounding autonomous weapon systems.
The International Committee of the Red Cross (ICRC), in its updated “Common Issues with Autonomous Weapon Systems” on October 8, 2026, pointed out that autonomous weapons may struggle to comply with the principles required by international humanitarian law, such as distinction, proportionality, and precautionary measures. These legal judgments must be based on specific circumstances during attacks, making it difficult to pre-program or simplify them into technical indicators.
Machines must also identify whether individuals are surrendering, injured, or protected by law, as well as whether the expected civilian harm is excessive. These judgments involve intent, behavior, and actual situations, which cannot be solely achieved through sensor data and program calculations.
ICRC warned that autonomous weapons may significantly accelerate target selection and engagement processes, heightening the risk of misjudgments, compressing human intervention and conflict mitigation space. When the use of force surpasses human judgment and control capacity, local engagements might escalate unpredictably, making it more challenging for conflicting parties to intervene in a timely manner, stop erroneous attacks, or de-escalate conflicts. The proliferation of autonomous weapons may also increase the risk of violating international humanitarian law.
In conclusion, autonomous lethal weapon systems have reached a crossroads where they may break free from ethical and legal constraints, leaving humanity with insufficient time, perhaps less than six months, to establish effective restrictions and retain substantial control.