Vietnam’s AI Ecosystem 2026

Artificial Intelligence (AI) has transitioned from an academic research topic to a national growth engine for Vietnam. In the 2024–2026 period, Vietnam’s AI ecosystem is developing on four parallel pillars: technology enterprises, research institutes and universities, the state’s policy framework, and human resources. This article outlines the overall picture, based on verified data sources.

1. Leading Enterprises

Private enterprises and state-owned corporations play a leading role, independently building computational infrastructure and “made in Vietnam” models.

FPT is the most valuable listed technology corporation on the Ho Chi Minh City stock exchange. In April 2024, FPT signed a Memorandum of Understanding with NVIDIA to build a $200 million “AI Factory,” operating as a sovereign cloud equipped with NVIDIA H100 GPUs and NVIDIA AI Enterprise software. FPT aims to expand its AI factories to Japan and South Korea.

Viettel — the military telecommunications group — approaches AI at a national infrastructure scale. Viettel AI has developed the Vietnamese language model VT-Super-120B-A12B (approximately 120 billion parameters), which is rated among the leaders in accuracy within its segment. Viettel Solutions also uses the NVIDIA NeMo Curator tool to clean Vietnamese data and train the Llama 3 ViettelSolution 8B model. In 2026, Viettel AI is confirmed to be building a national legal AI application on NVIDIA’s open model infrastructure.

VNG / Zalo owns Zalo — a messaging application used by approximately 85% of Vietnamese internet users. VNG has independently built its large language model KiLM from scratch; the 13B parameter version (late 2024) reportedly surpassed several international models in Vietnamese language processing within the VMLU evaluation framework. VNG also operates the cloud and AI platform GreenNode.

Vingroup once made significant bets on AI through VinAI Research (established 2019, creator of PhoGPT) and VinBigData (ViVi voice assistant in VinFast cars). However, this ecosystem is undergoing strong restructuring: in December 2024, Vingroup sold the medical AI startup VinBrain to NVIDIA; in April 2025, Qualcomm acquired VinAI’s generative AI division. This signals that Vietnam’s AI capabilities have become attractive enough for global corporations to acquire.

Beyond these “giants,” a layer of specialized companies and startups such as GreenNode (GreenMind inference model), MoMo (AI in fintech, payments), and numerous applied AI studios are shaping the middle tier of the ecosystem. Note: detailed financial figures for many of the aforementioned M&A deals have not yet been publicly disclosed.

2. Research Institutes and Universities

The academic foundation provides human resources and fundamental research. Notable centers include the Vietnam Institute for Advanced Study in Mathematics, AI groups at Hanoi University of Science and Technology, Vietnam National University, Hanoi, Vietnam National University, Ho Chi Minh City, and VinUni. The link between institutes and businesses is strengthening: in May 2025, VNG signed an agreement to sponsor 25 billion VND over 3 years for Vietnam National University, Ho Chi Minh City, committing to train at least 1,000 high-quality students.

The national strategy aims for at least one Vietnamese institution to be among the top 20 leading AI research and training institutes in ASEAN by 2030.

3. National Policy: Two Pillar Documents

Vietnam’s AI policy is guided by two key documents.

The National Strategy on Research, Development and Application of AI by 2030 (Decision 127/QD-TTg, issued January 26, 2021) sets objectives: to position Vietnam among the top 4 in ASEAN and top 50 globally in AI; to develop 10 reputable AI brands in the region; to establish 3 national AI innovation centers; and to build a legal framework, data infrastructure, and computational resources for AI.

Resolution 57-NQ/TW (signed December 22, 2024, by General Secretary To Lam) elevates science – technology, innovation, and digital transformation to a “breakthrough” status. The Resolution sets targets for the digital economy to contribute at least 30% of GDP by 2030 and 50% by 2045; to increase spending on science – technology to 2% of GDP within 5 years (from 0.82% in 2023); and to allocate at least 3% of the state budget for this task. On January 13, 2025, the Central Steering Committee for Science, Technology, Innovation, and Digital Transformation was established, directly led by the General Secretary — a rarely seen top-tier governance structure for the technology sector.

4. Human Resources: Opportunities and Bottlenecks

Human resources are both an advantage and the biggest bottleneck.

• Vietnam’s IT engineer workforce is estimated to exceed 530,000 people, with approximately 50,000–70,000 IT graduates annually.
• The number of specialized AI engineers is estimated at 18,000+ people — considered the largest in Southeast Asia, but still far short of demand.
• Demand for AI personnel is projected to increase by 74% in the 2025–2030 period.
• The IT industry faces a shortage of 150,000–200,000 programmers and engineers annually; approximately 45% of AI solution providers consider talent shortage their biggest operational challenge.

Key bottleneck: Vietnam excels at training junior and mid-level engineers, but severely lacks research leads, ML architects, and senior engineers with experience operating at production scale.

5. Investment and Infrastructure

Capital flow into AI is accelerating, primarily through two channels: domestic investment in infrastructure (FPT’s $200 million AI factory, Viettel’s and VNG’s data centers) and international M&A (NVIDIA acquiring VinBrain, Qualcomm acquiring VinAI’s AI division). In 2026, Vietnam became a focal point in NVIDIA’s “sovereign AI” strategy, with the announcement of a population-scale dataset developed in collaboration with FPT. Official comprehensive statistics on total AI investment across the entire industry are not yet publicly available.

Conclusion

Vietnam’s AI ecosystem in 2026 has moved beyond its nascent stage: it has enterprises independently mastering large models, a synchronized and decisive national policy, and large-scale GPU infrastructure. The remaining central challenges are high-level human resources and transforming technical capabilities into sustainable revenue-generating products. With a “sovereign AI” orientation and commitment at the highest leadership level, Vietnam is laying the foundation to achieve its goal of being among the top 4 in ASEAN by 2030.

References

• Decision 127/QD-TTg — National Strategy on AI by 2030 (LuatVietnam)
• Resolution 57 — Vietnam’s digital leap forward (VietnamPlus)
• FPT and NVIDIA ink MoU to build $200m AI factory (Vietnam News)
• Vietnam unveils first Vietnamese LLM on NVIDIA NIM (VietnamNet)
• Vietnam’s AI dreams run into a talent shortage wall (VnExpress)
• Qualcomm acquires generative AI division of Vietnamese startup VinAI (TechCrunch)
• Viettel develops 120-billion-parameter Vietnamese AI model (Vietnam.vn)