The war in Ukraine is seeing an increasing use of attack drones controlled by artificial intelligence (AI) to find, track, and strike targets with progressively less human assistance. While many drones in Ukraine are still remotely piloted, both Ukraine and Russia are rapidly adding computer vision, autonomous navigation, and AI-assisted targeting.

In July 2026, Ukrainian manufacturer Skyfall and U.S.-Swiss defense technology company Auterion began delivering 50,000 upgraded Shrike attack drones to Ukrainian forces. The drones incorporate Auterion’s Skynode S computer and Strike software, allowing an operator to identify a target before the drone takes over the final approach.

Using computer vision, those drones can continue tracking moving targets even when GPS signals or radio communications are jammed. The same hardware could eventually be upgraded to allow groups of drones to operate as coordinated swarms.

Russia and Ukraine are engaging in a new kind of arms race that could fundamentally change how future wars are fought. It is likely AI-controlled drones are already hitting targets in Ukraine, a first in the annals of modern warfare. There are several advantages and risks to engineering drones to be controlled by AI instead of human operators.

AI control solves one of the biggest problems drone operators have encountered in Ukraine. Electronic warfare systems can jam the communications links connecting conventional drones to their pilots. An AI-guided drone can receive instructions from a human, visually lock onto a vehicle or other target and complete the attack even after communications are disrupted. That means electronic jamming is less effective against AI drones.

Autonomy is another potential advantage. Ukraine’s Defense Ministry announced in June that Ukrainian developers had combat-tested an interceptor system in the Kharkiv region that automates about 95 percent of the process of attacking Russian Shahed-type drones. A human selects the target and authorizes the engagement, after which the interceptor can navigate toward, identify, track, and attack the incoming drone autonomously.

The scale of drone warfare helps explain why these developments matter. Ukraine’s Defense Ministry said in July that its unmanned forces had conducted more than one million verified drone strikes during 2026 and that drones were responsible for roughly 90 percent of attacks on Russian battlefield targets. Those figures are Ukrainian government estimates, but they illustrate how weapons once regarded as specialized equipment have become central to ordinary combat.

Russia is pursuing similar technologies. Russian forces have experimented with improved attack drones incorporating cameras, onboard computing, and anti-jamming equipment. In addition, Russian authorities said in July that they had intercepted AI-navigation-equipped Ukrainian drones allegedly intended for attacks on air bases deep inside Russia.

The long-term implications of AI-controlled drones could be global in scope. AI could allow one soldier eventually to supervise multiple drones instead of individually piloting one aircraft, enabling much larger attacks without proportionately larger numbers of personnel. Cheap autonomous drones might hunt artillery, armored vehicles, supply trucks, and even other drones, creating battlefields where exposed movement becomes extraordinarily dangerous.

Precision-guided weapons were once expensive systems available primarily to wealthy militaries. AI is beginning to put similar capabilities into drones for a fraction of the cost. If autonomous navigation, target recognition, and swarming become widespread, the defining weapon of future wars may not be tanks or fighter aircraft, but thousands of inexpensive intelligent machines capable of finding and attacking targets largely on their own.