Vietnam - Responsible AI Development (1290/QD-BKHCN)

Overview

Decision 1290/QD-BKHCN, issued by the Minister of Science and Technology on June 11, 2024, establishes the Guide on Principles for Researching and Developing Responsible AI Systems (Version 1.0). This document represents Vietnam's first comprehensive ethical framework specifically designed for artificial intelligence, establishing foundational principles for responsible AI management in the country. The Guide takes effect immediately upon issuance and is designed as a voluntary standard rather than mandatory regulation, reflecting Vietnam's balanced approach to AI governance.

Definitions

The Guide defines Responsible AI Systems as artificial intelligence technologies developed and deployed in accordance with ethical principles that balance innovation with human welfare. The framework encompasses AI systems across the development lifecycle, from research and design through deployment and operation. Human-centric AI is defined as systems that prioritize human well-being, dignity, and rights while delivering technological benefits. The Guide applies to all forms of AI systems including machine learning, automated decision-making systems, and related technologies used in Vietnam's digital ecosystem.

Governance and Institutional Framework

The Ministry of Science and Technology (MOST) serves as the issuing authority and oversight body for the Responsible AI Principles. MOST is responsible for updating the guidance as technology and international standards evolve. Because the instrument is established as a voluntary guideline, no administrative body enforces compliance or monitors self-regulation through formal sanctions. Organizations adopting the principles do so voluntarily, while sector-specific ministries may reference the guidance when formulating policy for their respective domains.

Key Focus Areas

• Cooperation: Encouraging collaboration among AI developers to ensure interoperability and compatibility between different AI systems.
• Innovation: Promoting the sharing of information and best practices while contributing to sustainable economic development.
• Transparency: Ensuring AI systems are explainable and their decision-making processes are understandable to users and affected parties.
• Controllability: Maintaining human oversight and the ability to intervene in AI system operations when necessary.
• Safety: Protecting users and society from harmful outputs and unintended consequences of AI systems.
• Security: Implementing cybersecurity measures to protect AI systems from malicious attacks and unauthorized access.
• Privacy: Protecting personal data processed by AI systems in compliance with data protection laws.
• Respect for Human Rights and Dignity: Ensuring AI systems do not discriminate or violate fundamental rights of individuals.
• User Support: Providing adequate assistance and resources to users of AI systems.
• Accountability: Establishing clear responsibility for AI system outcomes among developers, deployers, and users.

Implementation Framework

The Guide establishes a voluntary implementation framework that encourages organizations to adopt responsible AI principles through self-assessment and commitment. Organizations are encouraged to integrate the nine principles into their AI development processes, governance structures, and operational procedures. The framework supports flexible implementation allowing organizations to prioritize principles based on their specific AI applications and risk profiles. MOST provides guidance materials and support to assist organizations in implementing the principles. The voluntary approach is designed to encourage broad adoption while maintaining innovation flexibility.

Monitoring and Evaluation

As a voluntary standard, the Guide does not establish formal monitoring or enforcement mechanisms. Organizations adopting the principles are encouraged to conduct self-assessments of their AI systems against the nine fundamental principles. MOST may collect information on adoption rates and implementation practices to inform future updates to the Guide. The Ministry encourages stakeholder feedback on the effectiveness of the principles and areas requiring additional guidance. Industry associations and professional bodies may develop sector-specific implementation guidelines aligned with the national framework.

Penalties, Liability, and Appeals

The Guide on Responsible AI Principles is established as a voluntary standard and does not include specific penalties or enforcement provisions. Organizations that adopt the principles do so through voluntary commitment rather than legal obligation. However, violations of related mandatory laws (such as data protection, cybersecurity, and consumer protection laws) remain subject to existing legal penalties. The voluntary framework may inform future binding regulations, with the Law on Digital Technology Industry (2025) incorporating many of these principles into mandatory requirements. Organizations are encouraged to demonstrate compliance with the principles as a matter of best practice and corporate responsibility.

Relationship to Other Instruments

Decision 1290/QD-BKHCN implements and operationalizes principles outlined in the National AI Strategy (Decision 127/QD-TTg) which called for establishing legal frameworks for AI governance. The Guide complements existing laws including the Cybersecurity Law (2018) and data protection regulations. Many principles from the Guide have been incorporated into the binding requirements of the Law on Digital Technology Industry (2025), demonstrating the evolution from voluntary to mandatory standards in Vietnam's AI governance approach.

International Alignment

The nine principles established in Decision 1290/QD-BKHCN align closely with international AI ethics frameworks. The human-centric approach mirrors the OECD AI Principles adopted in 2019. The emphasis on transparency, accountability, and human rights reflects standards promoted by UNESCO's Recommendation on the Ethics of AI. The framework supports Vietnam's engagement with ASEAN digital cooperation initiatives and bilateral AI partnerships. By establishing voluntary principles aligned with international standards, Vietnam positions itself for harmonized AI governance across borders.

Implementation Timeline

Sources and References

Ministry of Science and Technology (MOST), Decision 1290/QĐ-BKHCN issued June 11, 2024.