How Russia Is Shaping AI on Its Own Terms
moskou, donderdag, 20 november 2025.
During the AI Journey 2025 conference in Moscow, President Putin personally showed interest in a human-like robot developed by Sberbank, powered by Russia’s homegrown AI model, GigaChat. But behind the technological showcase lies a grand strategic plan: Russia aims to develop its own sovereign AI models within ten years, while simultaneously tripling the electricity consumption of its data centres. The biggest surprise? Sberbank is set to lay off 20% of its staff using AI in the near future. What does this mean for the future of work and technology?
Russia’s AI Strategy: From Technological Showcase to National Ambitions
At the AI Journey 2025 conference held in Moscow from 19 to 21 November 2025, Russian President Vladimir Putin personally demonstrated interest in a human-like robot developed by Sberbank and equipped with the Russian large language model suite GigaChat [1][2][3]. This exhibition was more than a technological demonstration; it symbolised a large-scale national project to develop sovereign artificial intelligence. During the plenary session ‘Future with AI’, Putin emphasised that generative AI technologies are of strategic importance and that countries worldwide are competing to possess their own, nationally controlled language models [2][4]. Organised by Sberbank in collaboration with the Alliance for Artificial Intelligence, the conference attracted international experts from Russia, India, Brazil, the US, China, and other nations, with core themes including technological innovation, practical applications of AI in daily life, and the ethics of AI [1][2][3]. The planned rise of AI in Russia is supported by a national strategic plan aimed at contributing more than 11 trillion rubles to GDP by 2030 [2]. Within this framework, Russia seeks to build 38 nuclear power plants in Siberia, the Urals, and the Far East, with the goal of strengthening computing power for AI development while simultaneously tripling data centre electricity consumption over this decade [2][4].
AI in Practice: From Public Services to Personalised Communication
The integration of AI into the public sector goes beyond technological demonstrations. The Moscow Digital Healthcare platform has already been adopted by nearly 2,000 medical organisations across more than 70 regions, where AI assistants are used to analyse medical images and radiological examinations [2]. This application demonstrates how AI can process complex information and make it accessible to healthcare providers and patients alike. In the field of information dissemination and public communication, AI is being used to personalise content and reach target audiences more effectively. Sberbank announced that the GigaChat AI system will be deployed to handle questions during the direct line with Vladimir Putin, an annual government event allowing citizens to submit inquiries [4]. This illustrates how AI is being used as a tool for mass information exchange, enhancing the effectiveness of public campaigns by enabling rapid and scalable responses to widespread queries [2]. Additionally, experimental legal frameworks for AI are already being implemented in Moscow, Sakhalin, and other regions, with plans to expand them to a third of Russian territory, including the entire Far East [2]. These regions serve as pilot zones for managing AI in society, where transparency, accountability, and the ability to identify AI-generated content are central priorities [1][3].
The Human Cost of Automation: Sberbank and the 20% Layoff Initiative
One of the most controversial developments during the AI Journey 2025 was the statement by Sberbank CEO Herman Gref that the bank would lay off 20% of its employees using artificial intelligence, as part of a multi-agent AI system designed to detect ‘inefficient’ workers [5]. Gref stated that this reduction would be completed by 1 January 2026, following AI analysis that identified certain roles and projects as no longer efficient [5]. The layoff initiative, entirely driven by the need to free up resources for AI development investments, represents a direct response to a remark made by President Putin during the conference, in which he advised Sberbank to use AI to reduce ‘non-essential expenditures’ [5]. Gref defended the move by stating it was simply ‘a way to free up resources’ for innovation, although Putin himself rejected the terminology of ‘inefficient workers’, remarking that there are no ‘inefficient workers’, only ‘workers you have poorly guided’ [5]. Sberbank’s workforce, which stood at 294,578 full-time employees as of 30 September 2025, had declined from 308,092 employees at the end of 2024 [5]. AI investment costs at Sberbank are estimated at approximately one billion dollars per year, a figure Putin described as ‘an extremely expensive toy’, thereby increasing budgetary pressure [5].
Challenges Around Privacy, Inclusivity, and Reliability
Despite technological advances and the potential of AI in public information and communication, challenges around privacy, inclusivity, and reliability persist. The expansion of AI-based systems in public services, such as healthcare and policy communication, requires a strong foundation of transparency and accountability [1][3]. The Russian government aims to establish a national government framework for AI in the near term, with an operational headquarters to be created by the Presidential Administration and the government to oversee the implementation of generative AI across all sectors, regions, ministries, and agencies [2]. This centralised approach is intended to reduce the risk of fragmentation and unlawful applications. However, using AI to dismiss employees based on algorithmic evaluations raises ethical questions about the objectivity and fairness of these systems [5]. There is currently no public information available regarding the algorithmic criteria used to determine ‘inefficiency’, which increases the risk of unjust or discriminatory decisions [alert! ‘No information available on algorithmic validation or transparency in Sberbank’s layoff procedure’]. Furthermore, within the framework of experimental legal regimes for AI in Moscow and other regions, efforts are underway to draft regulations focused on identifying AI-generated content, a crucial element for maintaining public trust in information campaigns [2]. The future of AI in public communication depends on the ability to use this technology without undermining citizen trust.