Nornickel expects to earn minimum 50 billion rubles of income through use of AI in mining operations by 2030

MOSCOW. May 18 (Interfax) - Nornickel expects to earn a minimum of 50 billion rubles in income by 2030 from the use of artificial intelligence (AI) solutions in mining operations, CEO Vladimir Potanin said at a plenary session of the Digitalization of Industrial Russia 2026 conference.

The company uses AI to make decisions on crushing ore before flotation and production of concentrates, allowing the work to be optimized and for extraction to be increased by several percentage points. AI allows for decisions to be changed in a matter of seconds and thus makes for a perfect working regime; however, the human operator still remains central to the process, interfering when critical safety limits are exceeded.

"This is making a huge impact at our sites and our company is gaining around 10 billion rubles of additional profit from integrating this type of artificial intelligence," Potanin said.

"We are expecting to significantly increase it [the impact of using AI] in the near future, generating at least 50 billion rubles of economic benefits by 2030," he said.

Nornickel also uses the Magma system for planning and supervising mining operations, Potanin said. In the next few years, the company expects that the system will enable it not only to optimize management processes and logistics in the mining sector, but also to be able to carry out tasks which are beyond human ability. Potanin said that this referred to making swift changes to various production parameters: the drilling angle and the development of prepared ore reserves.

"We expect this to improve the quality of the ore by 5% and decrease the level of dilution and the extraction of superfluous rock from the mines," Potanin said.

He also said that it was especially important that access to various AI models be maintained and said that the comparison of these models allowed for the integration of this tool in Russia to be sped up. "With all the emphasis on Russian solutions - especially in the field of software - I would think that it is important to stay competitive and maintain access to different models at the stage of the development of artificial intelligence," he said. "I wouldn't want AI software to be more difficult to certify than software without AI."