AI Law 134/2025: a legal frame for enterprise AI in Vietnam

In 2025 Vietnam produced two texts touching AI: the Digital Technology Industry Law (71/2025) and now a dedicated act. On 10 December 2025, the National Assembly passed the AI Law (134/2025/QH15, effective 1 March 2026) — the first comprehensive AI law in Vietnam, with 8 chapters and 35 articles.

What the law sets out

• It classifies AI system risk into three tiers: high, medium and low — higher risk means stricter control.
• Transparency: AI systems that interact directly with people must be designed so users know they are dealing with a machine.
• AI-generated content must carry an identifying mark, to avoid misleading viewers.
• It bans using fakery or simulation of real people or events to deceive or manipulate perception and behavior.

What it means for enterprise AI

Together with the Digital Technology Industry Law (see Law 71/2025 on AI risk management), both point one way: enterprise AI must be transparent, governable and controlled by risk level. AI handling sensitive data deserves the most attention.

Why local AI is a compliance advantage

When AI runs on your infrastructure, data is not sent to outside services, and you control the whole processing lifecycle — easier to be transparent and traceable, in the spirit of the law. That is how Molly Reader works: reading and extracting document data with local AI.

In practice

The law is a frame; technical detail will emerge through guidance. But if you are bringing AI into operations, now is the time to design for transparency and control. For local document AI advice, book a consultation.