Vietnam Enforces Southeast Asia's First AI Law

Vietnam breaks new ground as Southeast Asia's first nation with a binding AI law

On 1 March 2026, Vietnam's Decree 13/2025 on artificial intelligence came into force, marking a pivotal moment in the region's approach to governing emerging technology. Vietnam is now the first country in Southeast Asia to enforce a comprehensive, standalone AI law, putting it ahead of far larger economies in the region and signalling that Hanoi is serious about shaping its digital future on its own terms.

The decree introduces a risk-based regulatory framework that draws clear inspiration from the EU AI Act, but is tailored to reflect Vietnam's particular development priorities, industrial structure, and social context. It is not a wholesale copy of European regulation. It is a deliberate adaptation, and that distinction matters enormously for how the region will watch and ultimately respond.

What the Law Actually Requires

Decree 13/2025 is built around several interlocking obligations that affect both domestic developers and foreign companies operating in Vietnam. The core requirement is mandatory registration of high-risk AI systems with the Ministry of Information and Communications, paired with impact assessments that must be completed before any such system can be deployed.

Transparency is a central pillar. High-risk AI systems must meet disclosure requirements, meaning users and affected parties must be informed when they are interacting with, or subject to, automated decision-making. The law also introduces strict restrictions on AI-generated deepfakes, reflecting growing concern about synthetic media's role in disinformation and fraud across the region.

"Vietnam's Decree 13/2025 establishes mandatory transparency requirements for high-risk AI systems, a national AI development fund, restrictions on AI-generated deepfakes, and data localisation rules for sensitive sectors." - Decree 13/2025 Summary, Vietnamese Ministry of Information and Communications

Beyond restrictions, the decree also creates infrastructure for growth. A national AI development fund is established under the law to channel resources into domestic AI research and capability building. This dual approach, regulating risks whilst investing in capacity, reflects a more sophisticated policy posture than simple prohibition.

Data Localisation and Sector-Specific Rules

One of the more commercially significant elements is the introduction of data localisation requirements for sensitive sectors. Companies handling AI-processed data in areas such as healthcare, finance, and national security will face requirements to store and process that data within Vietnamese borders. For multinational technology firms operating regional infrastructure, this creates real compliance complexity.

• Healthcare AI systems must comply with localisation rules for patient data
• Financial AI applications face both registration and data residency obligations
• National security-adjacent AI use cases are subject to the most stringent oversight
• Civil society organisations retain the right to seek redress through algorithmic transparency mechanisms

Industry Concerns and Civil Society Response

The reaction to Decree 13/2025 has been divided along predictable lines. Industry groups, particularly those representing startups and smaller technology companies, have raised concerns about the compliance burden. Registration requirements, impact assessments, and data localisation infrastructure all carry costs that established multinationals can absorb more easily than early-stage Vietnamese ventures.

Civil society organisations have taken a different view. Provisions on algorithmic transparency and citizen redress mechanisms have been broadly welcomed, particularly by digital rights advocates who argue that without such safeguards, AI systems in public services could entrench existing inequalities without any mechanism for challenge or correction. The tension between these two perspectives will define how the law evolves in practice.

"The law's provisions on algorithmic transparency and citizen redress mechanisms represent a meaningful step toward accountable AI deployment in a developing economy context." - Civil society assessment, Decree 13/2025 consultation process

The National AI Ethics Committee, a new body established under the decree, is tasked with reviewing emerging AI applications and providing policy recommendations as the technology evolves. This is a forward-looking mechanism that acknowledges the law cannot anticipate every development. Whether the committee has genuine independence and adequate resourcing will be critical to its credibility.

The Asia-Pacific Picture: Who Is Watching and Why

Vietnam's move does not exist in a vacuum. Across ASEAN, governments are actively developing their own AI governance positions, and Decree 13/2025 now serves as a concrete reference point for that work. Thailand, the Philippines, and Indonesia are all at various stages of drafting AI-specific regulation. Each has a different economic structure, regulatory tradition, and relationship with the major technology platforms operating in the region.

For context, Vietnam's 100-million-plus population and rapidly expanding technology sector give its regulatory choices outsized influence. A startup or platform that wants to access Vietnamese consumers must now comply with this framework. That creates de facto pressure on the entire regional ecosystem, not just companies operating in Vietnam specifically. This is sometimes called the Brussels Effect in reference to how EU regulation shapes global standards. Hanoi may be generating a similar, if more modest, dynamic within Southeast Asia.

China's own approach to AI regulation, including its rules on generative AI and algorithmic recommendations, has also shaped the broader regional conversation. You can read more about the trajectory of China's AI regulatory and investment strategy over the next five years for important context on how Beijing's approach compares with Hanoi's. Separately, the rise of competitive Chinese AI models, covered in our piece on free Chinese AI tools challenging GPT-5's dominance, underscores why governments across Asia are anxious to establish frameworks before the technology outpaces governance entirely.

Comparing Regional Approaches

The implications extend beyond compliance. As we have explored in our coverage of how people across Asia are actually using AI in their daily lives, adoption is accelerating at the grassroots level well ahead of formal governance. Laws like Decree 13/2025 are trying to catch up with behaviour that is already deeply embedded in commerce, communication, and public services.

The Deepfake Problem and Why Vietnam Moved Fast

One of the most politically sensitive provisions in the decree concerns AI-generated deepfakes. Vietnam, like many countries in the region, has experienced a surge in synthetic media used for fraud, political manipulation, and reputational damage. The law's restrictions on deepfakes are not merely symbolic. They reflect documented harms that have already reached Vietnamese courts and regulatory bodies.

This connects to a broader problem of AI-generated low-quality content flooding digital platforms across Asia. Our investigation into AI-generated content eroding social media quality across Asia provides wider context for why policymakers feel urgent pressure to act, rather than defer to self-regulation by platforms that have clear commercial incentives to allow such content.

1. Deepfake fraud cases involving financial scams have increased significantly across Southeast Asia in recent years
2. Political deepfakes targeting candidates and officials have been documented in multiple ASEAN elections
3. Platform self-regulation has proven insufficient, prompting legislative responses across the region

What Comes Next for Vietnam's AI Law

The enforcement phase is where the real test begins. A law that exists on paper but lacks the institutional capacity to enforce it provides only partial protection. Vietnam's Ministry of Information and Communications will need to build the technical expertise and staffing to evaluate AI impact assessments, maintain the high-risk system registry, and investigate complaints under the redress provisions.

The National AI Ethics Committee faces a similar challenge. Its credibility will depend on the independence of its members, the transparency of its deliberations, and whether its recommendations are actually incorporated into policy updates. These are institutional design questions that are often more consequential than the text of the law itself.

For companies operating in Vietnam or planning to enter the market, the practical priority is understanding which of their AI systems fall into the high-risk classification. The registration and impact assessment requirements create a compliance timeline that cannot be deferred indefinitely. Those who engage proactively with the Ministry of Information and Communications are likely to shape how implementation guidance develops.

Frequently Asked Questions

What is Vietnam's Decree 13/2025 and who does it apply to?

Decree 13/2025 is Vietnam's standalone AI regulation, in force from 1 March 2026. It applies to all developers and deployers of AI systems operating in Vietnam, including foreign companies. High-risk AI systems must be registered with the Ministry of Information and Communications and undergo impact assessments before deployment.

How does Vietnam's AI law compare to the EU AI Act?

Vietnam's decree draws on the EU AI Act's risk-based structure, including tiered obligations for high-risk systems and transparency requirements. Key differences include Vietnam's national AI development fund, its specific data localisation rules for sensitive sectors, and provisions tailored to the country's development stage and political context. Vietnam's law is binding and enforceable, not a voluntary framework.

Which other Southeast Asian countries are developing AI regulation?

Thailand, the Philippines, and Indonesia are all developing AI governance frameworks, but none has yet enacted a binding standalone AI law. Singapore has voluntary frameworks and sector-specific guidelines but has not moved to hard regulation. Vietnam's Decree 13/2025 is currently the only comprehensive, enforceable AI law in ASEAN.