Ministry of Economic Development of the Russian Federation

From the perspective of the Russian Federation, the success of the first Global Dialogue meeting would mean the shift from general discussion to a practical agenda. As a concrete outcome, it is important to confirm states' readiness to further elaborate the launch of a long-term education and science program in the field of artificial intelligence for developing countries. It is equally important to signal the need to reduce technological and expertise asymmetries, including through data sharing, access to libraries and repositories, joint research initiatives, and discussions on access to computational resources. Success would also include agreement on working principles: fair geographical representation, mandatory participation of specialists from developing countries, transparent management of conflicts of interest, and open publication of materials.

• Safe, secure and trustworthy AI
• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Transparency, accountability, and human oversight

The selection of these thematic areas is driven by the fact that they most comprehensively reflect the Russian priority, which is the launch of a long-term education and science program in the field of artificial intelligence for developing countries: • The "AI capacity-building" area is directly related to the need for workforce development, strengthening research competencies, and expanding states' access to knowledge, data, and computational resources. The key feature of the Russian initiative could lie in combining education with research activities: the program should facilitate the exchange of scientific data and datasets, provide access to open libraries and repositories, support joint competitive research challenges, offer small grants for publications and pilot projects, and, in the long term, provide arrangements on access to computational resources for research. At the same time, the program should be designed to prevent talent drain. It implies a combination of offline modules in Russia and online formats, dual affiliation for researchers, etc. • The areas concerning safe, secure, and trustworthy AI, as well as transparency, accountability, and human control, directly align with the concept of scientifically based expertise, independent risk assessment, and the preparation of comparable analytical materials for states. • The area addressing social, economic, ethical, cultural, linguistic, and technical implications of AI is important because AI governance cannot be reduced only to narrow security concerns. For the Russian Federation, it is essential to take into account the impact of AI on the labor market, education, quality of public administration, linguistic diversity, and developing countries' access to the benefits of technological progress.

The list of themes should be supplemented with issues that directly affect the feasibility of fair global AI governance: • Technological and expertise asymmetry between states, including unequal access to research centers, data, and resource bases; • Comparability of expert approaches: common methodologies, benchmark datasets, impact assessment templates, and other tools are needed to enable states to draw on a shared analytical foundation; • It is essential to address issues of linguistic and cultural diversity and prevent talent drain, so that capacity-building programs strengthen national scientific communities and help to address the priorities of researchers' countries of origin.

-

The Global Dialogue could serve as a universal platform capable of integrating discussions on AI governance principles with practical mechanisms for international cooperation. Its importance could lie in ensuring regular exchange of comparable information, building trust among states, and narrowing the gap between technologically advanced countries and those still developing their own AI policies. For the Russian Federation, it is particularly important that the Global Dialogue not be limited to declarative discussion, but instead facilitates the establishment of scientific expertise channels, the development of common methodological approaches, and the advancement of capacity-building programs.

It would be advisable for the Global Dialogue to build upon existing international and national mechanisms without duplicating their mandates. Key reference points could include the UNESCO Recommendation on the Ethics of Artificial Intelligence, the International Telecommunication Union's work under the AI for Good and the AI Skills Coalition initiatives, as well as science-focused initiatives implemented with the UN support.

Various stakeholders can make complementary contributions to the Global Dialogue's work. Governments set the political mandate and define societally significant priorities for regulation and development. The academic community can provide independent scientific expertise, develop assessment methodologies, and contribute to professional training. The private sector brings practical knowledge of technologies, infrastructure, and implementation practices, while international organizations can support coordination and scaling of successful solutions. Civil society is essential for ensuring that human rights, social, cultural, and linguistic dimensions are adequately reflected. It would be advisable to organize thematic working sessions and regional consultations between formal meetings of the Global Dialogue. To enhance effectiveness, the structure could include a dedicated scientific-expert track and a separate capacity-building track, along with a practice of preparing analytical briefs for each session. Such an approach would make the dialogue more substantive and less declarative.

In global discussions on AI governance, developing countries, small language communities, public universities, and research teams lacking access to large-scale computational resources remain significantly underrepresented. There are also insufficient practical requests from countries that are interested not only in risk management, but also in employing AI for development tasks such as education, healthcare, agriculture, public administration and scientific research. To address this imbalance, the following measures are needed: fair geographical representation; mandatory participation of specialists from developing countries in advisory mechanisms; open publication of materials; multilingual methodological foundation; and collaboration formats that enable researchers to maintain ties with their countries of origin.

The most useful formats would be those that combine political discussion with practical cooperation. These could include short, problem-oriented sessions on specific topics—such as model evaluation, computational resources, linguistic diversity, or workforce development—supported by pre-prepared analytical briefs. Another effective format could be joint meetings of government representatives and independent experts, where discussions are structured around comparable data and methodologies rather than solely around general political positions. For the education and science track, promising formats include project exchanges, network laboratories, joint competitive research challenges, register of available courses, benchmarks, and research partnerships.

Among useful practices, solutions that establish verifiable procedures for applying artificial intelligence deserve particular attention. In the Russian Federation, such an example is the integration of the updated National Strategy for Artificial Intelligence Development with coordination mechanisms and sectoral standardization; the Code of AI Ethics in the Field of AI, which facilitates the development of voluntary guidelines for technology implementation, plays an additional role. Another example from Russian practice is the AI Alliance Network, established in 2024 with Russia's participation. The Alliance unites research centers, industry associations, and technology organizations that develop AI based on principles of openness, cooperation, and responsibility. Its goal is to accelerate the development and deployment of AI technologies through joint competency-building, scientific collaboration, education, ethics and regulation in the field of AI, and the exchange of accumulated experience and best practices. Additionally, in 2025, the Russian Federation hosted the International Foresight - a strategic session on fundamental and exploratory research aimed at advancing AI. This initiative comprises sessions and in-depth interviews with scientists to identify AI long-term scientific priorities in advance. The Foresight engaged 270 researchers from 36 countries, who identified and updated 10 priority areas and 201 research tasks for the coming decade. Such forward-looking analysis enables researchers to focus on high-impact challenges. The Foresight report is publicly available worldwide in three languages, including English.