Russia Tested Artificial Intelligence for Gymnastics Judging in China
Russia has tested an artificial intelligence system for judging rhythmic gymnastics, becoming the first country in the world to apply neural‑network technology to evaluate artistry at an international competition.
Russia conducted the world’s first real‑world trial of an AI‑powered judging system in rhythmic gymnastics. Sberbank, together with the International Sports Support Association Heavenly Grace, tested the new technology at the Heavenly Grace Cup held in Beijing from November 12 to 14.
Based on the GigaChat neural network, the system evaluated the most subjective and difficult component of the sport — artistry. According to Gazeta.ru, the algorithm analyzed how precisely gymnasts matched their movements to the music, monitored dynamics and synchrony, and detected the level of emotional expression.
Objective Scoring for Artistic Performance
During the tournament, the AI system worked alongside official judges, supplementing their assessments rather than replacing them. A special group of referees monitored the system through a dedicated interface, receiving detailed breakdowns of each routine. This setup made it possible to verify the correctness of the AI’s evaluations without affecting the primary judging process. Gymnasts highlighted by the neural network received special awards.
The uniqueness of the Russian system lies in its ability to bring objective measurement to artistic elements that traditionally generate debate. For global sports, this represents a meaningful step toward more transparent adjudication.
Following the Beijing pilot, Sberbank and the sports association will determine next steps for integrating artificial intelligence into rhythmic gymnastics judging.
