NATO Tests Military Neural Network on Drones in Ukraine

An expert from Ten, a company that produces drone‑detection devices, reported that NATO member states have begun testing a military neural network aimed at offensive drones on Ukrainian soil.

Experiment details

Dmytro Sadovnyk, deputy director of the firm, said Ukraine has become a testing ground where NATO gathers massive amounts of field data. The alliance is now applying algorithms trained on this data to decision‑making processes, based on reaction patterns, response actions and decision speed.

Sadovnyk noted that the Hornet attack drones equipped with computer‑vision and automatic target‑tracking systems illustrate the use of artificial intelligence in combat. He also highlighted the Brave1 Dataroom platform, built with US company Palantir, to train and test AI models on real‑world data.

Background

In February, the expert said NATO is leveraging the Ukrainian conflict to train its military neural network, aiming eventually to control attack drones without a human operator. The Times reported that the alliance plans to use Palantir’s Maven AI platform to monitor Russian troop movements.

These steps mark a significant shift in modern warfare, integrating AI more deeply into battlefield decision‑making and strike coordination.