In brief
On 10 December 2025, the National Assembly of Vietnam passed the Law on Artificial Intelligence (AI) ("AI Law").
The AI Law is set to take effect on 1 March 2026, with grace periods for most existing AI system providers and deployers lasting until 1 March 2027 (or 1 September 2027 for AI systems in the fields of health, education and finance).
The AI Law will serve as a comprehensive legal framework for regulating AI system operations in Vietnam, replacing the entire general AI framework under the Law on Digital Technology Industry.
Although the AI Law contains many terms and references resembling those under the EU AI Act, the law contains many discrepancies and vague issues that are subject to further clarification from the Government.
Key takeaways
- Application scope: The AI Law applies to both domestic and foreign entities engaged in AI activities in Vietnam, covering research, development, provisions, deployment and use of AI systems.
- Definition of AI system: The AI Law defines an AI system as a machine-based system designed to implement AI capabilities with varying degrees of autonomy, capable of self-adapting once deployed. Based on explicitly or implicitly defined objectives, the system makes inferences from inputted data to generate outputs such as predictions, content, recommendations or decisions that may affect the physical or digital environment.
- Regulated roles in the AI value chain:
- Developer: An organization or individual that designs, builds, trains, tests or fine-tunes all or part of an AI model, algorithm or system, and has direct control over the technical methods, training data or model parameters
- Provider: An organization or individual that puts an AI system on the market or puts it into use under its name, brand or trademark, regardless of whether the system is developed by the provider themselves or by a third party
- Deployer: An organization or individual that uses the AI system under its control in professional activities, trade or service provision, excluding the case of its use for personal or non-commercial purposes
- User: An organization or individual that directly interacts with an AI system or uses the output of that system
- Risk classification and management: The AI Law adopts a risk-based approach to classifying AI systems, categorizing them into three levels of risk (i.e., high, medium and low) based on the level of impact on human rights, safety and security; fields where the system is used, especially essential fields or fields directly related to public interests; user scope; and scale of the impact of the system.
- High-risk AI (HRAI) systems will be subject to periodic audits or audits on signs of violation. The Prime Minister will issue the list of HRAI systems, which specifies those considered subject to pre-market conformity certifications and those that are not.
- Medium-risk AI systems will be supervised via reports, sample audits or assessments by independent organizations.
- Low-risk AI systems will be monitored and audited on incidents, complaints or when necessary to ensure safety.
- While there is no "unacceptable" risk category like the EU AI Act, the AI Law prescribes a list of prohibited acts, as set out below:
- Taking advantage of or appropriating AI systems to violate the law or infringe upon the legitimate rights and interests of organizations and/or individuals
- Developing, providing, deploying, or using AI systems for the prohibited purposes, notably using elements forging or simulating real persons or real events in order to intentionally and systematically deceive or manipulate human perception or behavior, causing serious harm to the rights and/or legitimate interests of individual; creating or disseminating forged content that is likely to cause serious harm to national security, public order, or social safety.
- Collecting, processing, or using data to develop, train, test, or operate AI systems in contravention of the laws on data, personal data protection, intellectual property and cybersecurity.
- Obstructing, disabling, or falsifying the mechanism of human supervision, intervention, and control of AI systems.
- Concealing information that is required to be disclosed, transparent, or explained; erasing and/or falsifying mandatory information, labels, and/or warnings in AI activities.
- Taking advantage of the research, testing, assessment, or verification of AI systems to commit acts contrary to the law.
- Local presence requirement for HRAI systems: The provider of an HRAI system subject to a pre-market conformity assessment must have a commercial presence or an authorized representative in Vietnam. For other HRAI systems, the provider only needs to have a lawful contact point in Vietnam. Details about the above obligations will be provided in the Government's guiding decree.
- Transparency requirements: The AI Law imposes different transparency requirements on the provider and the deployer. Generally, the provider must ensure that audio, image and video content generated by the AI system are marked in a machine-readable format per the Government's regulations, while the deployer must clearly indicate whether such content poses the risk of confusion regarding the authenticity of events and/or characters. Details about the transparency requirements will be provided in the Government's guiding decree.
- Timeline: The law generally enters into force on 1 March 2026, with the following grace periods granted to providers and the deployers of AI systems that have been put into operation before this effective date:
- 18 months for AI systems in the field of health care, education and finance
- 12 months for other AI systems
Outlook
As a framework law, the implementation of the AI Law will require the Government, Prime Minister and Ministry of Science and Technology to issue several implementing instruments. Several critical compliance components need to be clarified, such as detailed risk classification criteria, transparency and labeling requirements, incident reporting thresholds, conformity assessment procedures, and local presence obligations for foreign providers. Some draft documents are already being prepared for submission to the Government in February 2026.
Manh Hung Tran, Partner, Huu Tuan Nguyen and Huyen Minh Nguyen, Special Counsels, and Alex Do, IPTech Executive, have contributed to this legal update.
© 2026 BMVN International LLC. All rights reserved. BMVN International LLC is in strategic alliance with Baker & McKenzie (Vietnam) Ltd., a member firm of Baker & McKenzie International, a global law firm with member law firms around the world. In accordance with the common terminology used in professional service organizations, reference to a “partner” means a person who is a partner, or equivalent, in such a law firm. Similarly, reference to an “office” means an office of any such law firm. This may qualify as “Attorney Advertising” requiring notice in some jurisdictions. Prior results do not guarantee a similar outcome.