AI Adoption in Vietnam: Strategic Technology Priorities, the AI Law and Broad Risk Themes

1. Vietnam's AI Strategy: Strategic Technology Context under Decision 1131/QD-TTg

Vietnam is prioritizing the digital economy to drive its economic growth, positioning itself as a leading player in SEA. According to MIC (now the MST), Vietnam's digital economy is growing 20% annually, and is estimated to contribute 18.3% to GDP in 2024. By 2030, the digital economy is projected to contribute 30%–35% of GDP, with Vietnam aiming to enter the top 30 globally in digital economy contribution, up from 41st today.

• Resolution No. 57-NQ/TW: signed on 22 December 2024 and considered a breakthrough policy. Compared to Resolution No. 10, which focused on agriculture and poverty reduction 40 years ago, this resolution now extends to science, technology, innovation, and digital transformation.

• National Strategy on Research and Development and Application of AI (2021-2030): to position Vietnam as a global hub for innovation, AI solutions development, and applications by 2030.
• Industry momentum: local conglomerates such as Vingroup, Viettel, FPT, or VNG proactively establish partnerships with global leading tech companies such as NVIDIA, Google and Microsoft to build up infrastructure and AI talents in Vietnam. As labor costs trend higher, automation and IT investments are becoming more important for Vietnamese companies.
• Economic potential: Vietnam's economy is projected to gain up to USD120b in benefits from AI according to a report by the National Innovation Center on 12 June 2025. Vietnam's AI start-ups rank 2nd in SEA (US$780M raised until 2024).

Decision 1131/QD-TTg publishes the list of strategic technologies and strategic technology products, with an official count of 11 groups and 35 products. The AI-related group includes Vietnamese LLMs, virtual assistants, domain-specific AI, analytical AI, digital twins, and metaverse categories, indicating an intent to prioritize both foundational AI capability and applied AI products.

The same strategic list includes enabling layers such as cloud computing and large-scale data centers, robotics/automation, semiconductor chips (including AI chips), and next-generation mobile networks, which together suggest an integrated approach to industrial digital capability rather than "AI in isolation."

2. High-Level Legal Perimeter: Law on Artificial Intelligence (Law 134/2025/QH15) and adjacent rules

2.1 Who and what the AI Law covers

The AI Law regulates research, development, provision, deployment, and use of AI systems in Vietnam, and applies to both domestic and foreign organizations/individuals engaged in AI activities in Vietnam, with exclusions for AI activities exclusively serving national defense/security/cipher purposes. The law defines key lifecycle roles (developer, provider, deployer, user), which is relevant for allocating responsibilities across AI development and deployment chains.

2.2 Risk-based classification and high-risk AI systems

Practitioner summaries describe the AI Law as adopting a risk-based approach, typically classifying AI systems into high-risk, medium-risk, and low-risk categories based on their potential impact on human rights, safety and security, sector of application, scope and scale of users and impact, with increasing obligations for higher-risk systems (e.g., documentation, transparency, safety/oversight expectations).

Following the AI Law, on 30 June 2026, the Prime Minister issued Decision 33/2026/QD-TTg, effective 15 August 2026, establishing the list of high-risk AI systems. The list covers AI systems in six sectors: education, healthcare, banking, transport, ethnic affairs and religion, and judicial proceedings, and sets sector-specific criteria for determining risk levels.

The responsibility for classifying AI systems primarily rests with the provider and deployer (in case where modifications, integration, or changes in functionality create new or higher risks). This risk-based framing is consistent with Vietnam's stated objectives of enabling AI development while setting baseline safeguards.

2.3 Adjacent rules that shape AI deployment

• Law on Personal Data Protection No. 91/2025/QH15 (effective 01 January 2026): AI deployments involving personal data are materially affected by this framework, which applies broadly to personal data processing activities.
• Cybersecurity Law 2025: cybersecurity regulations may also shape architecture decisions for certain service providers.
• Law on Protection of Consumer Rights 2023: AI systems used in consumer-facing activities should take into account its requirements, including periodic assessment of the use of AI and automated systems, particularly for large digital platforms.
• Sector-specific laws: the application of AI in specific sectors must not only comply with the risk-management principles under the AI Law but also comply with other relevant sector-specific laws and regulations.

3. AI Adoption in Vietnam: Adoption Patterns, Use Cases and Challenges

According to BCG, AI is expected to significantly boost Vietnam's economy, especially in the manufacturing sector, which alone could see an impact of USD25–30b, followed by consumer products, financial services, telecom, and education. Many of these sectors have lagged in digitalization due to the challenge of processing large volumes of unstructured data and relying on outdated legacy IT systems. The new generation of AI, driven by large language models (LLMs), is tailor-made for these vertical markets.

Success in Vertical AI will be defined by its ability to tailor cutting-edge tools to the unique demands of each industry, transforming powerful yet generic AI into specialized, deployment-ready solutions that deliver tangible value.

Readiness

Vietnam currently ranks 45th out of 195 countries in the Government AI Readiness Index report by Oxford Insights (UK), surpassing the global average and maintaining its position regionally at 5th in Southeast Asia.

Private sector

AI has gradually expanded across industries in Vietnam since 2019, including banking, insurance, consumer finance, retail, healthcare, consumer packaged goods (CPG) and manufacturing. In the banking industry, one of the leading sectors in AI adoption, AI is primarily used for facial recognition, electronic know-your-customer (eKYC), optical character recognition (OCR), and call or voice bots deployed in contact centers, collections and credit card activation.

GenAI

Late 2023 and early 2024 marked a turning point for AI in Vietnam, driven by the rapid rise of the ChatGPT phenomenon. Some large companies have started piloting GenAI, mainly for internal purposes such as document and knowledge navigation, HR tasks like automated CV scanning and profiling, staff training, and IT support through code copilots. External-facing use cases are still largely focused on GenAI-powered chatbots and voice bots.

Value path

Deploy GenAI/AI in everyday tasks to realize 10% to 20% productivity potential; reshape critical functions for 30% to 50% enhancement in efficiency and effectiveness; invent new GenAI/AI business models to build a long-term competitive advantage.

Adoption scale

AI adoption in Vietnam surged in 2024, with 47,000 new enterprises implementing AI, bringing total adoption to nearly 170,000 firms, accounting for around 18% of all enterprises nationwide, and a year-on-year growth rate of approximately 39% in AI adoption according to a study by Amazon Web Services in 2025. AI solution provision is concentrated primarily in the IT sector (31%), followed by finance and banking (22%), education (17%), and healthcare and e-commerce (15%).

Structural challenges

According to the Vietnam Artificial Intelligence Annual Report 2025 by the Institute of Information Technology, Vietnam National University, based on a survey of nearly 500 enterprises and organizations conducted from July 2025, the most significant challenge facing AI solution providers is the shortage of high-quality talent (45%), while 23% struggle with data and computing infrastructure constraints.

Market structure

Vietnam's AI market is segmented between well-resourced technology conglomerates (FPT, Viettel, VNPT, and Vingroup), which lead in infrastructure and R&D but often keep applications within their own ecosystems, and emerging companies focused on specialized solutions.
Several Vietnamese players have gained global recognition: Abivin won the 2019 Startup World Cup for AI-powered logistics optimization, VinAI publishes in leading international journals, and NVIDIA's acquisition of VinBrain validated Vietnam's healthcare AI capabilities. Access for SMEs and the public sector remains limited.

4. AI Developers and Deployers in Vietnam (Selected Examples)

AI developers in Vietnam

AI deployers in Vietnam

5. AI Risks for Boards in Vietnam: Broad Risk Themes

For boards and management teams, the principal risks associated with GenAI and AI adoption in Vietnam extend beyond model performance. They arise from strategic readiness, implementation capability, technology infrastructure, and the evolving legal and governance environment. These risks should be assessed across the AI lifecycle and calibrated to the nature, scale, and potential impact of each use case.

5.1 Strategic and organizational risks

• Leadership capability and strategic direction. Limited AI proficiency at executive level can constrain informed decision-making, reduce awareness of enterprise-wide opportunities and risks, and leave initiatives without clear sponsorship.
• Governance and change management. Weak accountability, unclear decision rights, and resistance to new processes can prevent pilots from becoming sustainable operating capabilities.

5.2 Implementation and technology risks

• Data quality, availability, and lawful use. AI performance depends on sufficiently structured, reliable, representative, and legally usable data. Personal data requirements may restrict how employee, customer, and third-party information is collected, stored, transferred, and processed.
• Talent and operating capability. Shortages of experienced AI professionals at both leadership and operational levels can hinder solution design, vendor oversight, integration, validation, and ongoing monitoring.
• Infrastructure and scalability. Access to computing capacity, particularly GPUs, and the cost of building secure inference environments can be significant.
• Operational reliability. Model errors, system failures, or poorly integrated solutions can disrupt manufacturing, logistics, customer service, and other core processes.

5.3 Security, legal, and regulatory risks

• Security, confidentiality, and third-party exposure. The use of external models and service providers can create risks of data leakage, unauthorized access, and reuse of proprietary or personal information for model training.
• Transparency and human oversight. Organizations may be unable to explain how an AI system reached a decision or to identify when human review is required. This risk becomes more significant where outputs affect individuals, regulated activities, or business-critical decisions.
• Regulatory classification and compliance. Vietnam's risk-based AI framework requires providers and deployers to understand their roles, determine the applicable risk classification, and implement corresponding documentation, transparency, safety, and oversight measures.
• Cybersecurity and data architecture. Requirements affecting information-system security, data storage, and cross-border transfers may shape the choice between public cloud, private cloud, and on-premises deployment.

6. Conclusion

Taken together, these risks argue for a portfolio-based approach: prioritize use cases by value and risk, establish accountable governance, strengthen data and infrastructure foundations, and scale only where controls and operating readiness are proportionate to the intended deployment.

How can we help?

Contact our Legal and Financial Advisory teams to discuss how Vietnam's AI framework applies to your organization's AI strategy in Vietnam.