Russia unveils new AI training hub for combat drones

The new service gives developers access to a database of FPV battlefield footage, the military has announced The Russian Defense Ministry has revealed a new service providing drone developers with combat footage from FPV, multirotor, and fixed-wing drones to train military AI systems.

The ministry said the platform will help the armed forces adapt to evolving AI-related challenges.

Russian President Vladimir Putin said in September that Russia will continue to actively integrate digital technologies and AI-based solutions into its weapons development programs.

However, he warned of the risks of using AI in the military, noting that cases in which the technology makes decisions autonomously have been recorded.

The new service is protected by certified information-security systems, including modern encryption and authentication tools, the Defense Ministry said on Monday.

The platform also uses a standardized classification system for machine-learning data, allowing developers to organize and process information in a common format.

Developers upload the results of their work to the platform, where the data is checked for completeness and to ensure it follows the standardized format.

The Defense Ministry uses the data to build and train its own AI models.

More than 40 organizations involved in AI model training have joined the project, Deputy Defense Minister Dmitry Shcherbinin said.

Some are already integrating their trained models into combat drone systems, he added.

Read more WATCH Russian drone interceptor in action (MOD footage) In July, the ministry announced the development of a data repository for training neural networks.

Defense Minister Andrey Belousov said it will allow drone manufacturers to train neural networks for their aircraft and process data directly onboard.

He acknowledged the technical complexity of the work but said solutions are already being developed.

Belousov also ordered upgrades to a hardware and software system used to assess the performance of drone crews more accurately.