 |
| VinMotion’s humanoid robots at the national exhibition “80-Years Journey of Independence - Freedom - Happiness” in Hanoi__Photo: VNA |
Coming into force on March 1, the Artificial Intelligence Law (the Law) not only marks a significant milestone in the country’s institutional development for the digital economy, but also highlights the nation’s resolve to take the lead in mastering, governing and harnessing one of the foundational technologies of the new era.
As artificial intelligence (AI) and data emerge as two pillars of national digital transformation, the Law sets out fundamental principles governing AI activities in Vietnam. It stipulates that AI must serve human beings rather than replace human authority or responsibility. The development and application of AI must ensure safety, transparency and accountability, operate for the benefit of the people, and safeguard national interests as well as the lawful rights and interests of citizens.
At the same time, green, inclusive and sustainable AI will be promoted to support responsible technology development, stimulate innovation and broaden public access to AI applications.
With 35 articles arranged in eight chapters, the Law provides comprehensive regulations covering AI systems from research and development to the provision and use of such systems in Vietnam, except AI activities exclusively serving national defence, security and cryptography.
The Law applies to all Vietnamese agencies, organisations and individuals, as well as foreign organisations and individuals engaged in AI-related activities within Vietnam’s territory. It establishes a mechanism for governing cross-border activities, thereby affirming national sovereignty in cyberspace and the AI domain.
Safe and responsible AI governance
One of the notable contents of the Law is a risk-based management mechanism for AI systems. Under this mechanism, AI systems are categorised into three levels: high-, medium- and low-risk ones.
High-risk systems are defined as those likely to cause significant harm to life, health, or the lawful rights and interests of organisations and individuals, national interests, public interests or national security. Such systems are required to undergo conformity assessments, maintain technical documentation, keep operational logs, and ensure the possibility of human intervention.
Medium-risk systems must ensure transparency and provide explanations upon request of competent authorities, while low-risk systems are subject to regulatory oversight when signs of violations are detected.
By differentiating risk-based obligations, the mechanism avoids blanket restrictions and concentrates regulatory resources on areas with substantial impact.
The Law also places emphasis on ensuring the transparency of AI-generated contents, which is deemed as a highly practical issue, particularly as AI applications heighten the risks of misinformation, fraud and the widespread proliferation of deepfakes.
Accordingly, organisations and individuals providing AI-generated audio, image or video contents are required to label or otherwise mark such materials to enable users recognise their artificial origin. Furthermore, the deliberate and systematic use of fabricated or simulated elements of real persons or events to deceive or manipulate human perception or behaviour is prohibited.
Providers are also required to describe the intended use purpose, operational principles, data sources and risk-control measures of their AI systems. However, in order to safeguard technological confidentiality, the Law does not require the disclosure of source code, algorithms, parameters, or trade and technological secrets.
The Law also establishes a national AI ethics framework. Principles of safety, fairness, transparency and respect for human rights will be set out to serve as references for organisations and enterprises in developing and operating AI systems.
National AI infrastructure
The Law affirms that national AI infrastructure is strategic and put for investment by the State, enterprises and social organisations. National AI infrastructure will be developed as a unified, open and secure ecosystem, designed to enable connectivity, interoperability and scalability, facilitating the AI development and application.
The Law also encourages organisations and individuals to make their AI-related data resources available to state agencies and other entities based on agreement.
Noteworthily, the Prime Minister will promulgate a list of datasets serving AI development in essential sectors. Priority will be given to cultural data, Vietnamese language and minority ethnic language data, as well as data relating to administrative procedures, healthcare, education, agriculture, environment, transport, socio-economic development, and other key fields.
The Law further provides for the establishment of a single-window AI portal, together with a national database on AI systems. These digital platforms are expected to facilitate the receipt and registration of participation in regulatory sandbox mechanisms, while disclosing information on AI systems in Vietnam.
Promotion of innovation and access to AI
The Law underscores the State’s policy of promoting AI technology groups aligned with Vietnam’s domestic context. The national AI strategy will be reviewed and updated at least once every three years, or upon the occurrence of significant technological or market changes.
 |
| Students experience operating robots at the workshop “Applying AI in Children’s Learning” in Da Nang city__Photo: VNA |
Organisations and individuals operating within the AI ecosystem will enjoy the highest incentives and support under the laws on science and technology, investment, digital technology industry, high technology and digital transformation. They will also be assisted in accessing infrastructure, data and testing environments for the research, production and commercialisation of AI products and services.
The State will support the development of the AI ecosystem and market by prioritising the use of AI products and services in accordance with the law on bidding. Furthermore, the Government will establish a national AI development fund to mobilise, coordinate and allocate resources for the promotion of AI research, development, application and governance.
The development of AI human resources will be carried out comprehensively at all educational and training levels. Fundamental AI contents, computational thinking, digital skills and technology ethics will be incorporated in compulsory general education curricular. Vocational education and higher education institutions are encouraged to develop training programmes in AI, data science and related disciplines.
The Law specifies that policies will be adopted to facilitate public access to, and enjoyment of the benefits of, AI. It also promotes the development and application of AI in support of social welfare, as well as the preservation, protection and advancement of national cultural identity.- (VLLF)
