Russia - AI Ethics Guidelines (2021)

Overview

The Code of Ethics in the Field of Artificial Intelligence (the "Code") is a voluntary national ethical framework that was presented and signed at the I International Forum "Ethics of Artificial Intelligence: The Beginning of Trust" in October 2021. It was adopted on 2021-10-26 and is intended to guide developers, providers, users, research organisations and public authorities engaged with AI systems in the Russian Federation. The Code sets high-level ethical principles and recommends institutional mechanisms and operational practices to promote the responsible design, deployment and use of AI systems in civil (non‑military) domains. Primary public information and commentary on the Code and related implementation activities are available via official government portals (for example, the national AI portal: https://ai.gov.ru/ai/regulatory/?utm_source=openai) and official agency announcements (for example, Rospatent notices: https://rospatent.gov.ru/ru/news/rospatent-prisoedinilsya-k-kodeksu-etiki-v-sfere-is-22122022?utm_source=openai). The full text of the Code as published by the AI Alliance is additionally available at https://a-ai.ru/ethics_EN/index.html.

Definitions

The Code adopts working definitions to clarify the scope and applicability of its provisions. It defines key terms such as "systems based on artificial intelligence", "developers", "users" and "stakeholders in the AI lifecycle", and uses these definitions to delimit responsibilities, recommended processes and the stages of development, deployment and operation covered by the Code. These working definitions are intended to make clear which actors and which types of AI activity are expected to consider the Code's principles and to assist organisations in determining how to implement the Code’s recommendations in practice. Authoritative descriptions and the Code text are discussed on official portals (see https://ai.gov.ru/ai/regulatory/?utm_source=openai).

Governance and Institutional Framework

The Code envisages governance and institutional mechanisms to support its voluntary implementation across sectors. It recommends the establishment of coordinating bodies (for example, a national commission or coordinating body on AI ethics), the appointment of ethics officers or ombudsmen within organisations, the possibility of sectoral ethics committees and the creation of collective registries or registers of trustworthy developers. Following adoption, a national commission for implementation was established at the national level to coordinate rollout and related activities. The Code anticipates cooperation among public authorities, developers, research organisations and users to operationalise ethical principles and to develop methodological recommendations and best practices; official descriptions of these governance arrangements and the accession of federal agencies are reflected on government portals (see https://ai.gov.ru/ai/regulatory/?utm_source=openai and https://rospatent.gov.ru/ru/news/rospatent-prisoedinilsya-k-kodeksu-etiki-v-sfere-is-22122022?utm_source=openai).

Key Focus Areas

• Human‑centredness and protection of human dignity and autonomy: the Code emphasises respect for human rights and human autonomy in the design and deployment of AI.
• Non‑discrimination and fairness: the Code requires attention to risks of bias and discrimination and recommends measures to prevent unfair outcomes.
• Risk orientation and precaution: the Code promotes risk‑based assessment and the application of the precautionary principle where harms are plausible but uncertain.
• Responsibility and accountability: actors are encouraged to take responsibility for foreseeable impacts of AI systems and to adopt internal accountability mechanisms.
• Transparency and informed interaction: the Code recommends transparency toward affected persons, including identifying instances where people interact with AI systems.
• Data protection and security: obligations include ensuring lawful processing of personal data and appropriate data‑security measures.
• Operational measures and governance: the Code advises appointment of ethics officers/ombudsmen, creation of sectoral committees, maintenance of internal ethical charters and adoption of procedures for risk and humanitarian‑impact assessments.
• Trust‑building and voluntary adherence: the Code functions as a soft‑law instrument intended to build public trust through voluntary commitments, public signing and peer oversight rather than statutory compulsion.

Implementation Framework

Implementation of the Code is based on voluntary organisational commitment. Signatory organisations publicly declare adherence by signing and committing to follow the Code’s provisions; implementation relies on internal policies such as ethical charters, documented procedures for risk assessment and mitigation, and methodological recommendations produced by coordinating bodies. Recommended institutional steps include the appointment of ethics officers or ombudsmen within organisations, establishment of sectoral ethics committees where appropriate, development of internal processes for identification and labelling of AI interactions with people, and maintenance of registers or directories of trustworthy developers. The Code also encourages the development and periodic update of best‑practice guidance and methodological materials to support consistent internal implementation across sectors. Government portals and agency notices summarise these implementation expectations and report on events and signings (see https://ai.gov.ru/ai/regulatory/?utm_source=openai and https://rospatent.gov.ru/ru/news/rospatent-prisoedinilsya-k-kodeksu-etiki-v-sfere-is-22122022?utm_source=openai). The full text published by stakeholders is available at https://a-ai.ru/ethics_EN/index.html.

Monitoring and Evaluation

Because the Code is a voluntary instrument, monitoring and evaluation are primarily based on non‑statutory mechanisms: public disclosure of commitments, peer review, sectoral signings and reputation‑based incentives. The Code recommends regular internal reviews of organisational practices, updates to procedures and adoption of lessons learned. National coordination arrangements, including the commission established to oversee implementation, facilitate monitoring through reporting of signatories, publication of sector‑specific guidance and the development of methodological recommendations for risk and humanitarian‑impact assessments. Official portals track accession by public bodies and the publication of related materials (see https://ai.gov.ru/ai/regulatory/?utm_source=openai and https://rospatent.gov.ru/ru/news/rospatent-prisoedinilsya-k-kodeksu-etiki-v-sfere-is-22122022?utm_source=openai).

Penalties, Liability, and Appeals

The Code does not create statutory sanctions, administrative penalties or a formal appeals mechanism. Enforcement is reputational and cooperative: adherence is supported through voluntary commitment, public declarations by signatories, peer mechanisms and sectoral oversight. The Code is designed to inform future regulatory and standardisation work rather than to establish new penal regimes; any questions of legal liability arising from AI use remain subject to applicable statutory law and sectoral regulation rather than to the Code itself. Official commentary and agency communications reiterate the voluntary, soft‑law nature of the instrument (see https://ai.gov.ru/ai/regulatory/?utm_source=openai and https://rospatent.gov.ru/ru/news/rospatent-prisoedinilsya-k-kodeksu-etiki-v-sfere-is-22122022?utm_source=openai).

Relationship to Other Instruments

The Code is positioned as complementary to Russia’s National AI Strategy and to sectoral regulation. It is intended to inform standardisation, sectoral policies and possible future statutory instruments. Authorities and stakeholders have discussed ways to incorporate ethical provisions into legislation and sectoral rules while preserving the Code’s voluntary, trust‑building character. The Code serves as a reference point for regulators, standards bodies and industry stakeholders seeking to align sectoral standards and voluntary conformity schemes with ethical principles endorsed at the national level. Official portal materials describe the Code’s role in the broader regulatory and policy landscape (see https://ai.gov.ru/ai/regulatory/?utm_source=openai).

International Alignment

The primary focus of the Code is national: it establishes a Russian voluntary ethical framework for AI and serves as a basis for national implementation and sectoral guidance. The published materials and official descriptions emphasise its national, voluntary character and do not present it as a binding instrument of international law. The Code may nonetheless be used as reference material in discussions with international standardisation bodies and multilateral fora, but such international engagements and alignments are not specified as formal obligations within the Code itself. For official summaries and announcements concerning the Code’s adoption and national implementation, see https://ai.gov.ru/ai/regulatory/?utm_source=openai and related agency communications at https://rospatent.gov.ru/ru/news/rospatent-prisoedinilsya-k-kodeksu-etiki-v-sfere-is-22122022?utm_source=openai.

Implementation Timeline

Sources and References