What happened
Vietnam's Deputy Prime Minister Ho Quoc Dung signed Decision No. 33/2026/QD-TTg, dated June 30, 2026 and published July 3, 2026, promulgating an official List of High-Risk Artificial Intelligence Systems spanning six sectors: education, ethnicity and religion, healthcare, banking, litigation (justice), and transportation. The decision also stipulates operating principles and a compliance roadmap for providers and deployers of these systems. It takes effect August 15, 2026, operationalizing obligations under Vietnam's broader AI/data-cybersecurity legal framework (including the 2025 Law on Cybersecurity, which took effect July 1, 2026).
Why it matters
This is a binding government decision (not a consultation) that formally classifies specific AI system categories as high-risk, triggering compliance, oversight, and likely licensing/reporting obligations for AI providers and deployers operating in banking, healthcare, education, justice, and transport in Vietnam. It sets a concrete enforcement baseline in one of ASEAN's stricter emerging AI/data regimes and signals the operationalization (from principle to enforceable list) of Vietnam's AI governance framework.
Action needed
Organizations developing or deploying AI systems in the six listed sectors serving the Vietnamese market should map their systems against the official high-risk list, and prepare compliance measures ahead of the August 15, 2026 effective date.