Vietnam Enforces New AI Law Mandating Content Labeling and Risk Classification

Vietnam has officially implemented Law No. 134/2025/QH15, a landmark piece of legislation that establishes a comprehensive framework for artificial intelligence. Effective as of March 1, 2026, the law requires all AI-generated images, audio, and video to be clearly watermarked or labeled. This move aims to ensure transparency and prevent the spread of misinformation through deepfakes and manipulated media. ### Risk-Based Oversight. The legislation introduces a risk-based classification system similar to international standards like the EU AI Act. Systems are categorized into high, medium, and low risk levels, with high-risk applications in healthcare and finance facing the strictest oversight. Developers of high-risk systems must undergo mandatory conformity assessments and register with a national database to ensure safety. ### Global Compliance Standards. Beyond labeling, the law imposes significant requirements on both local and foreign AI providers. Foreign entities offering high-risk AI services are now required to establish a commercial presence or appoint a representative within Vietnam. This ensures that regulators have a direct point of contact for compliance and incident reporting. ### Prohibited AI Practices. To safeguard social order, the law strictly prohibits certain AI applications such as social scoring and subliminal manipulation. The use of AI for illegal surveillance in public spaces or the creation of deceptive deepfakes for fraud is also banned. These protections are designed to prioritize human rights over unrestricted technological expansion. ### Economic Incentives. While the law introduces strict guardrails, it also offers incentives to foster a thriving digital economy. Provisions include the creation of a National AI Development Fund and regulatory sandboxes for testing new technologies. These measures support startups by providing tax breaks and a controlled environment for experimentation. ### A Growing International Trend. Vietnam joins a growing list of nations implementing structured AI governance to mitigate emerging digital threats. As countries like China and those within the European Union refine their own rules, Vietnam’s approach signals a regional shift toward mandatory transparency. Global tech companies must now navigate a patchwork of local laws that demand higher levels of accountability.