Vietnam Law on Artificial Intelligence → 适用法域 → Vietnam
AI 来源复核 · Jev 5%1 条出处记录 · 1 条报告快照
· Global AI Regulation Monitor · 全球 AI 监管追踪
独立 AI 来源复核 · 2026-11-01
越南法律3月1日生效,存量系统过渡期不改变in_force,律师事务所analysis分类正确。
Codex 独立来源复核 · Jev 5% · 不计人工核验逐字原文依据
comprehensive Vietnam's Law on Artificial Intelligence, passed by the National Assembly on 10 December 2025 and in force from 1 March 2026, is the country's first comprehensive, standalone AI statute and the world's third national horizontal AI law after the EU AI Act and South Korea's AI Framework Act. It sits alongside the Law on Digital Technology Industry (No. 71/2025/QH15), whose AI provisions continue to apply. The Law reaches Vietnamese and foreign organisations and individuals engaged in the research, development, provision, deployment or use of AI systems in Vietnam. Activities serving only national defence, security and cryptography are out of scope. It distinguishes four roles — developer, supplier, implementing party and user — and allocates duties among them, including serious-malfunction obligations to rectify, suspend or recall a system and notify the competent authority. Transparency: suppliers must ensure a system interacting directly with humans is designed so the user knows it is AI, and that generated audio, image and video content is tagged in machine-readable form. Implementing parties must notify the public where AI-generated or AI-edited content could confuse audiences about the authenticity of events or people, and must label simulations of real persons or events. Prohibited acts apply irrespective of risk tier: deceptive manipulation, exploitation of vulnerable groups, dissemination of false content seriously threatening national security or public order, unlawful data use, obstruction of required human oversight, and concealment or falsification of mandatory disclosure or labelling. Penalties run from administrative sanctions to criminal liability, with a civil-liability regime placing compensation on the implementing party for damage caused by a compliant high-risk system, subject to recourse and to force-majeure and claimant-fault defences. 10 Dec 2025 — National Assembly passes the Law on Artificial Intelligence Vietnam adopted a risk-based, EU-inspired horizontal AI statute, moving in a single step from strategy-and-guidance to comprehensive legislation. The Law sets out national policy on AI as a driver of growth and innovation, and establishes the state management architecture for AI activities. Allen & Gledhill — Vietnam's new Law on Artificial Intelligence: risk-based regulatory framework in force 1 March 2026 (30 Mar 2026) <https://www.allenandgledhill.com/vn/publication/articles/32667/s-new-law-on-artificial-intelligence-risk-based-regulatory-framework-in-force-1-march-2026> Duane Morris — Vietnam: the first Law on Artificial Intelligence (3 Mar 2026) <https://blogs.duanemorris.com/vietnam/2026/03/03/vietnam-the-first-law-on-artificial-intelligence-what-you-must-know/> 1 Mar 2026 — Law enters into force with staged grace periods AI systems already operational at commencement may continue to run during transition: 18 months in healthcare, education and finance, and 12 months in all other fields. The grace periods are not absolute — where the competent authority finds a serious risk of harm it may order temporary suspension or termination regardless. Allen & Gledhill — Vietnam's new Law on Artificial Intelligence: risk-based regulatory framework in force 1 March 2026 (30 Mar 2026) <https://www.allenandgledhill.com/vn/publication/articles/32667/s-new-law-on-artificial-intelligence-risk-based-regulatory-framework-in-force-1-march-2026>
越南 AI Law 自 2026 年 3 月 1 日生效,存量系统分 12/18 个月过渡。