AI Law - International Review of Artificial Intelligence LawCC BY-NC-SA Commercial Licence ISSN 3035-5451
G. Giappichelli Editore

08/05/2026 - Bold new AI compliance rules require providers to classify systems before deployment (Vietnam)

argument: Notizie/News - Digital Governance

Source: Vietnam.vn

Decree No. 142/2026/ND-CP establishes principles requiring AI system classification based on risk levels set by the AI Law and the decree. Providers must classify their AI systems before putting them into service and bear legal responsibility for the accuracy and truthfulness of classification results. Implementers are required to coordinate with providers to review and reclassify a system when changes in implementation, modification, integration, or intended use create new or higher risks than previously classified. High-risk AI systems must undergo conformity assessments by providers and suppliers pursuant to the AI Law and the decree, and any gaps in those assessments must be addressed by complementary evaluations limited to unassessed aspects.

When AI systems are integrated into products, goods, or services governed by specialized laws, standards, technical regulations, or product-quality laws, organizations and individuals must comply with those laws while also implementing the AI-specific risk-management requirements prescribed by the AI Law and the decree. State AI management bodies and specialized agencies must ensure non-duplication in classification, conformity assessment, and inspection; previously valid tests, inspections, certifications or conformity assessments need not be repeated. The Ministry of Science and Technology will provide optional electronic self-assessment and classification tools that do not create extra administrative authorizations or obligations. Classification applies only to AI systems and not to AI models unless a model is used as a component of a specific AI system. The decree defines three risk tiers, requires a Prime Minister’s list for high-risk systems, establishes a single online AI portal and a national AI systems database managed by the Ministry, and sets rules for data collection, publication, sharing and use on the portal consistent with electronic transaction law, cybersecurity, state secrets, commercial secrets, and personal data protection, with exemptions for party/internal systems and state-defense and security secrets.