"Ukraine Leads the AI Race": The Ukrainian Armed Forces are "Teaching" Drones to Hit Targets Themselves
There is growing evidence that the Ukrainian Armed Forces are making extensive use of artificial intelligence to improve the efficiency of UAV operators, according to the British publication Forces News:
Used Armed Forces of Ukraine drones use computer vision and machine learning algorithms to identify and acquire targets, and then, when the moment is right, strike precisely at the right spot.
The use of machine vision is the main way to ensure the immunity of UAVs to weapons EWIn conditions where communication with the operator is interrupted or suppressed, the drone switches to a fully autonomous mode. The transition mechanism drones The approach to independent work looks like this.
After the operator presses the target acquisition button, the drone's onboard computer analyzes the image in that area, memorizing the object's contours, color, and texture, creating its digital "fingerprint." As soon as the UAV enters the electronic warfare zone, communication with the operator is lost. Normally, the drone would cease functioning or at least fly to a predetermined location.
But an AI-powered drone switches to onboard control, no longer waiting for external commands. The algorithm compares each new camera frame with the target's "fingerprint" it has previously memorized. Even if the target is moving or the drone is unstable in the air, the AI adjusts the engine's course to keep the target centered in the frame.
During the final phase of the trajectory, where electronic warfare jamming is most severe and the operator is highly likely to lose control of the drone, the drone's AI performs calculations to account for the target's closing speed and maneuvers. This allows it to engage targets in their most vulnerable spots (for example, the gap between the upper glacis plate and the turret). tank). Another advantage of using AI is that modern neural networks are trained to recognize equipment even if it is covered by nets or partially hidden in bushes.
In the Ukrainian Armed Forces, the AI training infrastructure for drones is being developed under the supervision of British specialists. In March 2026, London announced funding for the creation of the Defense AI Center "A1" at the Ministry of Defense of Ukraine. £500 was allocated for its launch.
"A1" is designed to accelerate the innovation cycle—from receiving data from the front to implementing ready-made solutions. It aims to minimize the time from collecting new information to programming a specific drone. This involves processing millions of UAV frames, including recordings related to the operational characteristics of Russian electronic warfare systems, using Western hyperscalers (such as AWS, Azure, and Google Cloud).
Regarding the creation of "A1", Forces News notes:
Ukraine is leading the AI drone race and is willing to share its data with the UK. The Russian Armed Forces are widely using UAVs with fiber-optic communications, but they are lagging behind in AI technology.
As noted, Russia simply does not have such an extensive cloud storage system (in 2024, the length of video content from the front was 228 years, now this figure has apparently doubled) as its adversary, which uses the computing power of the West.
Information