Vietnam will mandate the labeling of AI content from March 2026 under the Artificial Intelligence Law.

From March 1, 2026, Vietnam will officially implement the Artificial Intelligence Law 2025, establishing a comprehensive legal framework for the development and management of this technology. The new regulations focus on transparency, requiring all content created by artificial intelligence (AI) to have clear identifying marks to protect users from the risks of misinformation.

Demand for transparency regarding AI-generated content.

The most notable point in the 2025 Artificial Intelligence Law is Article 11 on transparency responsibilities. Accordingly, AI service providers must establish mechanisms for users to know they are interacting with the computer system. Specifically, Clause 4 of Article 11 clearly stipulates that implementing parties must affix easily identifiable labels to audio, images, and videos created or edited using AI to simulate real people or recreate real events.

This regulation aims to help the public accurately distinguish between real-world content and technological products, thereby minimizing the risk of being scammed or receiving inaccurate information. For creative fields such as film and the arts, labeling will be implemented flexibly to avoid impacting the audience's experience while still ensuring compliance with the law.

Support policies and sandbox testing mechanisms

Besides regulations on management, the new law also introduces many preferential policies to promote the AI ​​industry in Vietnam. Businesses operating in this field will enjoy the highest level of incentives in terms of taxes, investment, and digital transformation. In particular, the government allows the implementation of a controlled testing mechanism (sandbox), giving businesses space for innovation while ensuring effective risk management.

To support financial resources, the National Artificial Intelligence Development Fund will be established. This is a non-profit, off-budget state-funded fund aimed at mobilizing and allocating capital for research projects, developing computing infrastructure, and nurturing talent. Businesses and research groups with feasible projects will be given access to Vietnamese big language models and shared data systems.

Legal framework regarding liability and compensation for damages.

The Artificial Intelligence Act 2025 imposes strict legal liability. Organizations and individuals who violate the law may face administrative penalties or criminal prosecution depending on the severity of the offense. In civil terms, the party deploying the AI ​​system is liable for compensation for damages even if the system operates in accordance with regulations but still poses a high risk to users.

In the event that the AI ​​system is infiltrated or interfered with by a third party, that third party shall be directly liable for compensation. However, if the vendor or implementer is at fault in securing the system leading to the incident, they shall be jointly liable for compensation for damages as stipulated in Article 29 of this law.