Vietnam Passes First AI Law: What Legal Teams Need to Know
Hanoi - December 11, 2025. Vietnam's 15th National Assembly has approved the country's first Law on Artificial Intelligence, alongside amendments to the Law on Intellectual Property and a revised Law on High Technology.
The AI Law spans eight chapters and 35 articles. The two companion laws include targeted IP amendments (three articles) and a refreshed High Technology framework (six chapters, 27 articles). All three measures passed with strong majorities during the Assembly's 10th session.
Core Approach and Policy Direction
Presenting the report on the draft AI Law, Minister of Science and Technology Nguyễn Mạnh Hùng noted broad consensus on its urgency. The law sets out core principles, prohibited acts, and a risk-based management framework.
The model draws on global practice: regulate inputs through data governance, govern use through legal and ethical standards, and manage outcomes through accountability and redress. This structure gives regulators multiple points of control across the AI lifecycle.
Risk-Based Oversight, With a Moving "High-Risk" List
High-risk AI systems face tighter obligations, drawing lessons from the EU and the Republic of Korea. Rather than freezing categories in statute, the law allows the Prime Minister to issue and continuously update the list of high-risk AI systems in "real time" (Clause 4, Article 13).
This design choice reduces the need for legislative amendments as technology and use-cases shift. For context on comparable models, see the EU AI Act.
Standards and Conformity Assessment
Conformity assessment applies only to systems designated as high-risk on the Prime Minister-approved list. Technical standards will be the reference point for assessment, allowing lower-risk systems to operate without mandatory conformity processes unless later designated.
Governance: Centralized Oversight
Proposals for an independent National AI Committee were dropped. Oversight will be centralized under the Government, with the Ministry of Science and Technology serving as lead coordinator across ministries and sectors.
Development Measures: Sandboxes, Funding, and Incentives
To balance control with growth, the law introduces controlled sandbox testing that can include partial or full exemptions from specific compliance requirements. A National AI Development Fund will support projects through special financial mechanisms.
Startups will have access to a voucher scheme to offset early costs. The incentive design takes cues from Japan's pro-innovation approach while maintaining firm guardrails for higher-risk deployments.
Effective Date and Timeline
The Law on AI takes effect on March 1, 2026. Secondary regulations, technical standards, and implementing guidance are expected to clarify obligations ahead of that date.
Practical Next Steps for Counsel and Compliance Teams
- Inventory current and planned AI systems across the business; flag use-cases that could plausibly be designated high-risk.
- Map data inputs and data lineage for each system; tighten data governance, access controls, and audit trails.
- Review vendor and model-provider contracts for audit rights, transparency, quality metrics, incident reporting, and indemnities tied to AI risk.
- Establish internal risk classification and human oversight procedures now, so they can align quickly with the Prime Minister's high-risk list once issued.
- Prepare technical documentation templates (intended purpose, training data sources, evaluation results, known limitations, change logs).
- Set up monitoring for updates to the high-risk list and related implementing decrees/circulars from the Ministry of Science and Technology.
- Evaluate whether sandbox participation could reduce time-to-deployment for novel use-cases.
- Coordinate with IP counsel to assess how the IP amendments may affect model training inputs, licensing, and derivative outputs.
- For firms in designated sectors, review overlaps with the revised Law on High Technology and any sectoral rules that may reference it.
What to Watch
- Initial issuance of the Prime Minister's high-risk list and the cadence for updates.
- Accreditation criteria and scope for conformity assessment bodies.
- Sector-specific guidance (finance, health, public services) and any cross-border applicability rules.
- Details of the National AI Development Fund, eligibility thresholds, and the startup voucher mechanism.
Context: Legislative Package
The Assembly advanced three related measures in one session: the AI Law, focused on risk and governance; targeted IP amendments; and an updated High Technology Law. Together, they set a clearer track for compliance, innovation, and enforcement in AI-intensive industries.
If your team needs structured upskilling on AI governance and risk, see role-based programs at Complete AI Training.
Your membership also unlocks:
