Law on Artificial Intelligence

Jurisdiction:

Vietnam

enforcing

Effective:

Mar 1, 2026

Authority:

Ministry of Science and Technology

Official text Verified Mar 26, 2026

Obligations Covered

Risk Assessment Human Oversight Transparency & Disclosure

Risk-Based AI Classification #

Obligation:

Risk Assessment

enforcing

Effective:

Mar 1, 2026

Risk tier:

high-risk

Scope:

providers, deployers

Requirements

Requirement	Details
Risk classification	AI systems classified and managed based on risk levels
High-risk list	Prime Minister prescribes the list of high-risk AI systems
Conformity assessment	High-risk systems require conformity assessment and certification before use
Grace periods	18 months for legacy systems in health/education/finance; 12 months for others

Penalties

Violation	Fine
Non-compliance	Per inspection and examination provisions (Chapter VI)

Human Oversight and Ethical Principles #

Obligation:

Human Oversight

enforcing

Effective:

Mar 1, 2026

Risk tier:

all

Scope:

providers, deployers

Requirements

Requirement	Details
Human-centric	AI activities must place humans at the center
Human authority	AI must not replace human authority in decision-making
Monitoring protection	Prohibits obstructing or falsifying human monitoring, intervention, and control
Ethical principles	Transparency, fairness, non-bias requirements

Penalties

Violation	Fine
Non-compliance	Per inspection and examination provisions (Chapter VI)

AI Content Labeling and Disclosure #

Obligation:

Transparency

enforcing

Effective:

Mar 1, 2026

Risk tier:

high-risk

Scope:

providers

Requirements

Requirement	Details
Conformity information	Must publish conformity information for high-risk systems
Labeling requirements	Prohibits concealing, erasing, or falsifying mandatory labels and warnings
Deepfake prohibition	Prohibited to use deepfakes to deceive people

Penalties

Violation	Fine
Non-compliance	Per inspection and examination provisions (Chapter VI)