Vietnam released Decree 142 on 7 May 2026, finalising the implementation rules for its landmark AI Law that took effect on 1 March. The decree clarifies the country's approach to AI risk classification, content labelling, and — most consequentially for operating companies — incident reporting timelines. Serious AI incidents must now be reported within 72 hours of discovery, with a five-working-day window for lesser incidents. The decree retroactively took effect from 1 May, giving businesses minimal lead time to comply. The move makes Vietnam the first country in Southeast Asia to pair a binding AI statute with a fully operational enforcement decree, an important marker for a region where most governance frameworks remain voluntary.

Decree 142 ties the formal list of high-risk AI systems to a prime ministerial decision still under preparation, meaning companies must begin risk self-classification while awaiting official categorisation. Scale of impact remains the primary classification criterion. The decree also mandates clear labelling of AI-generated audio, image, and video content, though it leaves specific display formats to deployers and delegates detailed technical guidance to the Ministry of Science and Technology. Foreign providers of high-risk AI systems — those requiring pre-market conformity assessments — must establish a commercial presence or appoint an authorised representative in Vietnam, a condition that will materially affect deployment decisions by global AI firms.

The decree's significance extends well beyond Vietnam's 100-million-person market. Analysts and legal observers have noted that Vietnam's AI Law, enacted in December 2025, mirrors the EU AI Act's risk-based structure and could catalyse regulatory tightening across Southeast Asia. With Indonesia finalising a Presidential AI regulation, Malaysia exploring dedicated AI legislation, and Thailand revising its draft AI principles, the arrival of concrete enforcement mechanics in Vietnam puts pressure on neighbouring governments to accelerate their own timelines. For regional technology operators and investors, Decree 142 is the first proof that Southeast Asia's paper commitments to responsible AI are acquiring legal teeth.

Australia's AI governance posture stands in sharp contrast to Vietnam's Decree 142. The National AI Plan, confirmed in December 2025, explicitly stepped back from mandatory guardrails in favour of existing technology-neutral laws and a new AI Safety Institute — promised AUD 29.9 million to launch in early 2026. The Australian Competition and Consumer Commission has prioritised anti-AI-washing enforcement in its 2026–27 agenda, and Privacy Act amendments requiring automated decision-making disclosure take effect in December 2026. But Australia has no sector-spanning incident reporting mandate or content labelling requirement to match Vietnam's. For Australian firms deploying AI in Vietnam — including financial services and healthtech operators — Decree 142 creates immediate, non-negotiable local compliance obligations that sit well above anything currently required at home.

Vietnam's Enforceable AI Rules Highlight Australia's Compliance Gap

India published national AI Governance Guidelines under MeitY in November 2025, establishing principles around trust, accountability, and responsible innovation — but with no binding obligations and no enforcement mechanism. Vietnam's Decree 142, by contrast, creates hard legal deadlines, mandatory labelling, and potential criminal liability for non-compliance. India's Digital Personal Data Protection Act creates binding data obligations but lacks AI-specific provisions. As the Indian government evaluates whether to legislate beyond guidelines, Vietnam's experience — moving from law to enforcement decree in under six months — offers both a model and a warning: speed can compress consultation periods in ways that leave businesses scrambling. Indian AI exporters selling into Vietnam will face local representative requirements for high-risk systems, adding cost and complexity to their Southeast Asian expansion.

India Watches Vietnam's AI Decree as Its Own Governance Framework Remains Voluntary

Singapore remains ASEAN's most sophisticated AI governance jurisdiction — its Model AI Governance Framework, AI Verify toolkit, and MAS FEAT Principles form the region's most coherent voluntary architecture. But Vietnam's Decree 142 raises a pointed question: how long can a voluntary model hold its regional credibility as binding obligations proliferate elsewhere? Singapore's IMDA and PDPC co-produced the world's first AI governance testing toolkit, and the city-state is consistently ranked among Asia's top AI-ready governments. However, sector-specific guidance without cross-economy incident reporting requirements means that a serious AI failure in, say, financial services may face very different scrutiny in Singapore than the same system operating in Hanoi. Executives running dual-market operations must now manage materially different compliance clocks.

Singapore's Voluntary Model Faces New Pressure as Vietnam Mandates AI Incident Reports

vietnam

southeast-asia

indonesia

malaysia

thailand

ministry-of-science-and-technology

Vietnam's new Decree 142 requires AI providers to report serious incidents within 72 hours and mandates labelling of AI-generated content — here's what it means for regional operators.

72-hour AI incident rule: Vietnam's Decree 142 explained

Decree 142 finalises enforcement rules for Vietnam's AI Law, creating the region's most operationally specific compliance obligations for foreign tech companies.

For companies operating across Southeast Asia, Vietnam's Decree 142 is the clearest signal yet that the region's AI governance posture is shifting from aspiration to enforcement. The 72-hour serious incident reporting requirement, content labelling mandates, and local-presence obligations for foreign high-risk AI providers are operationally demanding in ways that no ASEAN-level guideline has been. Legal advisers at Baker McKenzie and Duane Morris have flagged that several critical compliance components — including detailed risk classification criteria and conformity assessment procedures — still await the Prime Minister's formal high-risk system list, meaning companies must act under ambiguity.

The broader ASEAN picture remains uneven. Singapore leads on governance toolkits but has no binding AI statute. Malaysia is positioning itself as an AI Nation by 2030 but has yet to legislate. Indonesia's Presidential AI regulation is expected but unsigned. Vietnam's move — accelerated from concept to enforcement decree in under six months — has compressed the timeline for regulators across the bloc. The ASEAN Digital Economy Framework Agreement, expected to be signed by end of 2026, may formalise interoperability standards for AI governance across member states, making alignment with Vietnam's framework now strategically sensible rather than merely obligatory.

Vietnam's Decree 142 creates immediate obligations that will reverberate across ASEAN's still-voluntary governance landscape.

Southeast Asia's First Enforceable AI Decree: What Regional Operators Must Do Now

Vietnam's Binding AI Decree Sets 72-Hour Incident Rule for Southeast Asia

Vietnam's Law on Artificial Intelligence officially takes effect March 1, 2026, making it the first Southeast Asian nation to establish a comprehensive, risk-based AI regulatory framework. The National Assembly passed the legislation in December 2025 with 90.7% approval, positioning Vietnam among global early adopters alongside South Korea, which launched its AI Basic Act in January 2026. The law applies extraterritorially to any foreign organization whose AI systems impact Vietnamese users, markets, or national interests, requiring local representation for high-risk applications.

The legislation mirrors the EU AI Act's risk-based approach but focuses on innovation support rather than restrictive compliance. AI systems are classified into three risk tiers, with high-risk applications in finance, healthcare, and education requiring pre-market conformity assessments, ongoing risk management, and mandatory registration through a national AI portal. Administrative fines reach VNĐ2 billion ($75,800) for organizations, with serious violations potentially incurring penalties up to 2% of annual revenue.

Beyond regulation, Vietnam introduces innovation incentives including regulatory sandboxes for AI testing, a National AI Development Fund, and voucher schemes for startups. Companies operating AI systems before March 1 receive transition periods: 18 months for finance, healthcare, and education sectors, and 12 months for other industries. The Ministry of Science and Technology leads centralized oversight, with detailed implementation guidelines expected in coming months.

Australia is expected to firm up its AI guardrails by 2026, lagging behind Vietnam and South Korea's enforcement timelines. The Australian government has indicated it may adopt a risk-based approach similar to the EU AI Act, following global regulatory alignment trends that previously influenced data privacy laws. Australian businesses operating in Vietnam face immediate compliance obligations under the new law's extraterritorial provisions, potentially serving as a compliance testing ground for future domestic regulations.

Australia Watches as Vietnam, Korea Sprint Ahead on AI Laws

India's Digital Personal Data Protection Act enforcement scheduled for mid-to-late 2025 creates new obligations for AI systems processing personal data, but lacks comprehensive AI-specific guidance. India has adopted a pro-innovation stance developing policies that mobilize workforce engagement in AI use and development, taking a risk-based approach rather than broad regulations that could stifle innovation. Vietnamese success with balanced regulation-innovation frameworks may accelerate India's move toward comprehensive AI legislation, particularly given growing policy coordination discussions between the nations.

India's AI Strategy Under Pressure as Vietnam Takes Lead

Singapore continues betting on voluntary frameworks like AI Verify and the Model AI Governance Framework, maintaining its light-touch regulatory approach while neighboring countries implement binding legislation. Singapore's mature governance regime anchors regional standards, with financial institutions facing binding obligations under MAS guidelines requiring governance-by-design for credit, insurance, and wealth management AI systems. The city-state's toolkit approach provides flexibility but creates competitive pressure as Vietnam offers similar sandbox incentives with clearer regulatory certainty.

Singapore's Voluntary AI Framework Faces New Regional Competition

south-korea

Vietnam's groundbreaking AI law creates compliance obligations for foreign firms, risk-based controls, and sandbox incentives starting March 1.

Vietnam AI Law March 2026: First Southeast Asia enforcement

Comprehensive regulation with extraterritorial reach takes effect March 2026, following South Korea's January launch.

South Korea's AI Basic Act took effect January 22, 2026, followed by Vietnam's AI Law on March 1, establishing the world's first comprehensive national AI regulatory frameworks in active enforcement. The rapid rollout marks APAC's transition from legislative drafting to real-world compliance obligations, with businesses across tech, fintech, healthcare, and manufacturing facing urgent adaptation requirements.

The fragmented regulatory landscape creates cross-jurisdictional compliance challenges, with Singapore maintaining voluntary frameworks, China implementing content regulations, and Australia expected to firm up guardrails by 2026. Industry experts note growing policy coordination between Singapore, India, and South Korea to harmonize ASEAN standards, focusing on risk management, explainability, and algorithmic accountability across sectors.

Arc Brief

Ministry Of Science And Technology

News Coverage

Vietnam's Binding AI Decree Sets 72-Hour Incident Rule for Southeast Asia

Vietnam Makes History as First Southeast Asian Nation to Enforce AI Law

Vietnam Makes History as First Southeast Asian Nation to Enforce AI Law

Vietnam Makes History as First Southeast Asian Nation to Enforce AI Law