On October 3, 2025, the Moscow Institute of Physics and Technology (MIPT) hosted the section "Application of AI Technologies and Systems in Shaping Industry 5.0", organized by the Club of Directors of Science and Innovation (iR&Dclub) and Rostec State Corporation. The event was attended by representatives of science, industry, government corporations and the legal community. The section featured seven speakers whose reports covered the legal, technological and strategic aspects of the introduction of artificial intelligence into industry and the scientific and technical sphere.
The legal part of the section was opened by Alexey Darkov, Adviser and head of the Intellectual property practice at VERBA LEGAL. He analyzed in detail the legal status of artificial intelligence as software and an object of intellectual property. The issue of the legality of datasets has become more acute: in most jurisdictions, including Russia, the consent of copyright holders is required for their use, with the exception of certain cases provided for by law, such as in the EU. "When developing your own AI, you need to settle legal relations with the authors of the materials, or make sure that the acquired dataset is legally clean," Alexey Darkov emphasized.
Anastasia Myrsina, Senior Associate at VERBA LEGAL, continued the topic of AI regulation by presenting a comparative analysis of the approaches of leading jurisdictions. The European Union has a risk-based approach: regulation depends on the level of potential harm, not on the technology itself. The US is dominated by a fragmented and industry-specific approach with an emphasis on self-regulation and standards such as NIST AI RMF. China uses a centralized model: AI systems are pre-tested for compliance with ideological and social norms. In Russia, according to the speaker, a hybrid system is being formed based on the national program "Digital Economy" and the Strategy for the development of AI until 2030, supported by a series of GOST standards on the lifecycle, trust and architecture of AI systems. Anastasia Myrsina stressed the need for companies to audit their AI processes right now, evaluate work with personal data, and implement managed AI use systems, including internal local regulations and the allocation of roles of responsible persons.
Anatoly Antipov, Director of NTI and Analytics at INVENTORUS, uses an intelligent platform as a tool for science-based decision-making in the field of high-tech innovations. INVENTORUS collects and structures information from open sources — patents, scientific publications, R&D, dissertations — forming a single body of scientific and technical knowledge (INVENTOCORE). In May 2025, the platform opened free access to the database and was recognized as the most advanced and promising NTI platform in Russia as part of a study by the RUSSOFT Association. In July, the INVENTORUS platform, the intellectual core of which is the AI assistant Nikola, became a finalist in the third season of the Know Our Growing Russian Brands competition. As part of the finalists' exhibition, Russian President Vladimir Putin personally got acquainted with the platform. He praised the efforts of the team of creators, noting: "The main thing is that this is no longer reverse engineering, but products created on the basis of their own developments. We especially want to support such brands." The President instructed the Ministry of Education and Science to include INVENTORUS in the national subscription so that all Russian universities would have access to it, which has been done since October.
Andrey Baden, the founder of the Product Lab, presented the concept of the X10 Products Laboratory as a response to the challenges of Industry 5.0. In the X10 Laboratory, generative AI is used at all stages of the product lifecycle — from niche search and idea generation to positioning, building growth funnels and scaling. An AI mentor helps teams conduct client research, formulate value propositions, test hypotheses, and build roadmaps. At the same time, the person remains in the center.: AI accelerates experiments and improves the quality of solutions, but it does not replace strategic thinking. Baden emphasized that the combination of Product Focus methodology and generative AI makes it possible to shorten the cycle "from idea to prototype" from a certain number of years or months to a matter of weeks and create personalized, breakthrough products.
Rustem Valiev, PhD in Economics, CEO of Fortek LLC, in his speech emphasized the fundamental difference between modern neural network models and classical expert systems. According to him, generative AI, despite its impressive expressiveness, remains a "black box": it does not explain why it makes a particular decision, and its conclusions cannot be verified logically. In contrast, expert systems based on explicitly defined rules and knowledge bases ensure transparency, reproducibility, and trust, qualities that are critical for high—tech industries such as engine manufacturing, energy, or defense. Valiev stressed that the future of the 5.0 Industry does not lie in blind trust in LLM, but in hybrid architectures, where generative AI can generate hypotheses, but the final verification, logical conclusion and decision—making remain with expert systems based on proven knowledge and formalized rules.
Alexander Reinhardt from a technology company Arrein.com (Almaty) proposed the concept of "Industry 5.5" as a transitional stage between the human-centered 5.0 Industry and the hybrid intelligence of the future. The speaker emphasized the fundamental difference between intelligence (the ability to process information) and reason (the ability to create meaning): AI has the former, but not the latter, so it remains a tool, not a subject. In Industry 5.5, according to his model, innovations must undergo a "stress test" for sustainability using environmental simulation, digital twins and expert systems in order to survive in real conditions.
Evgeny Pavlov, Head of the Department of Innovative Development of JSC UEC, completed the section. He stressed that the UEC is choosing a hybrid AI implementation model: a combination of competence centers and decentralized implementation, which allows flexible adaptation of technologies to the needs of different departments.
In conclusion, the participants agreed that generative AI has not yet been implemented into real production chains — most of the cases are pilot or demonstration in nature. The main barriers are the lack of standardization, low reliability of models, poor integration with industrial systems, and, most importantly, the unavailability of the organizational environment. At the same time, traditional AI solutions - expert systems, big data analysis, specialized scientific and technical solutions support services — are already bringing measurable benefits. All the speakers emphasized that AI is not a substitute for humans, but a reinforcement tool. The sustainability of innovation depends not so much on technology as on the human factor, organizational culture, and system architecture. Such platforms as the MIPT section contribute not only to the exchange of experience, but also to the development of specific solutions aimed at increasing the technological sovereignty and competitiveness of Russian industry in the context of global digital transformation.