Deputy director of the drone-detection company Tenʹ, Dmitry Sadovnik, believes NATO is using the conflict in Ukraine as a testing ground for its own military neural network algorithms. In his view, in the near future there will be AI systems on the battlefield that will not need a human operator at all.

Sadovnik described the current situation as a dangerous trend, arguing that a global training process for neural networks is underway and that such systems cannot exist without large datasets. He said this link between data and algorithms is inseparable and that, in his assessment, the collective NATO bloc is training its military neural network directly in wartime conditions.

He noted that the massive collection of data and the testing of new types of weapons in Ukraine allow the West to prepare for a possible third world war. According to him, the top priority now is precisely the training of neural network algorithms.

Sadovnik stated that on the battlefield, in real time, various patterns of action and counteraction against drone swarms are being worked out. He said the behaviour of targets is being analysed, including how they try to evade the impact of electronic warfare systems, and added that from the hardware point of view «something changes there every day».

The report recalls that earlier Google, OpenAI, Anthropic and xAI received contracts from the Pentagon worth up to 200 million dollars each. These agreements are aimed at integrating advanced AI systems into the work of the US military department.