30

On June 20, 2025, a Seminar "Forecasting and Management" was held in the Academician Inozemtsev conference hall (3rd floor).

Oleg Rogov, Head of the Reliable and Secure Intelligent Systems Group of the Artificial Intelligence Research Institute (AIRI), Head of the Joint AIRI-MTUCI "Safe AI Laboratory", Cand. of Science (Physics and Mathematics), presented the report "Generative AI. New Opportunities and Challenges of Global Security".

The era of fundamental language (LLM) and multimodal (MMLM) models opens up new opportunities for socio-economic development and scientific progress, but at the same time carries systemic risks. This report presented a comprehensive analysis of the key aspects of this era, based on four interrelated research areas:

• - global competition and assessment of its dynamics;
• - approaches to assessing model reasoning chains and creating specialized benchmarks of complex scenarios with manual and automatic validation to ensure reliability;
• - research in the field of generative content moderation, digital watermarks and methodologies for countering multidimensional cyber threats;
• - aspects and methods for minimizing bias to increase the objectivity of the result.