The legislation, which took effect on Sunday, introduces a risk-tiered model where AI providers – both local organisations and foreign entities with a presence in the country – must classify their systems as low, medium or high risk based on guidelines from the Ministry of Science and Technology.
Companies must also explicitly label AI-generated content such as deepfakes and disclose to customers whether they are interacting with an AI bot instead of a human agent.
Advertisement
Vietnam’s National Assembly passed the law in December. The approach, which resembles the European Union’s AI Act, places a strong emphasis on accountability, transparency and safety.
The Southeast Asian nation joins a small cohort, including the EU and South Korea, that have enacted legislation for AI, even as countries around the world move to implement guidelines and regulations on generative AI use.
The law “paves the way for Vietnam to deeply integrate with international standards while maintaining digital sovereignty”, the government said in a December report.