The Vietnam AI Economy 2025 report, released in June by the National Innovation Center (NIC), the Japan International Cooperation Agency (JICA), and the Boston Consulting Group (BCG), offered an ambitious projection: by 2040, AI could contribute up to $130 billion to Vietnam’s GDP, equivalent to around 25 per cent of the country’s current economic scale. This figure is striking not only because of its magnitude but also due to the strategic vision it reflects, where AI is no longer viewed as a distant frontier but as a central pillar in the country’s development strategy.

However, behind these headline-grabbing numbers lies a far more complex reality. Vietnam’s AI ecosystem remains in its formative stage. Current AI applications, from chatbots to text recognition, are still largely limited to pilot or demonstrative phases. The integration of AI into key sectors such as manufacturing, finance, agriculture, and healthcare remains scattered and lacks depth.

In this context, the question is no longer “Is AI important?” but rather “Does Vietnam have the capacity to turn AI into a real engine of growth? Or will it fall into the all-too-familiar cycle of past tech booms, such as blockchain, Industry 4.0, and smart cities, which made waves for a few years before fading into obscurity?” The answer does not lie in policy declarations, but in the level of organization and the ability to mobilize resources in a sustainable way.

AI and the future of Vietnam’s economy

Unlike previous industrial revolutions, which relied on physical resources or cheap labor, AI places the global economy in a new paradigm, with competition based on data, algorithms, and information processing capabilities. AI is no longer a standalone technology sector but a foundational layer capable of reshaping every field, from manufacturing and smart agriculture to public governance and national security.

Vietnam is betting on AI as a leapfrogging strategy to climb the global value chain. This motivation is driven by both internal and external factors. Internally, Vietnam’s economy has reached the limits of its labor and resource-intensive growth model. Productivity remains low compared to its regional peers. According to a 2023 report from the Asian Development Bank (ADB), Vietnam’s workplace productivity stood at only 7.6 per cent of Singapore’s, 18 per cent of Malaysia’s, and 45 per cent of Thailand’s. Meanwhile, its young workforce is now shrinking, and labor costs are rising. These challenges demand a qualitative breakthrough rather than mere quantitative expansion, and AI is seen as the guiding tool.

From a global perspective, Singapore has successfully embedded AI into every level of national governance, from transportation and healthcare to public services. The government not only launched its national AI strategy early, in 2019, but also established a binding AI Governance Framework, attracting substantial investment into AI services. China, meanwhile, has adopted a “large-scale AI” model with massive public funding for surveillance, facial recognition, and State-serving supercomputers. Meanwhile, the US relies on the private sector, led by global giants like OpenAI, Google DeepMind, and Nvidia, as well as top-tier universities such as MIT, Stanford, and others, to provide the talent infrastructure.

Vietnam now faces a complex challenge: how to build an AI ecosystem that is open enough to attract private investment, sovereign enough to control its data, and responsible enough to avoid technological dependence on foreign powers. This is not just a technological issue, it is a national development strategy in the post-industrial era.

2020 is considered the official starting point of the National AI Strategy. By 2024, Vietnam had made rapid progress in terms of quantity, but not yet in quality. According to the Vietnam AI Economy 2025 report, Vietnam’s AI market reached approximately $750 million last year, with an annual growth rate of 15-18 per cent; higher than the global average of around 12 per cent. The fastest-growing AI applications are in fintech, followed by online education, healthcare, and e-commerce.

In terms of startups, data from PitchBook Data Inc. shows that Vietnam currently has 765 AI and machine learning startups. These account for about one-quarter of all tech startups in the country. Vietnam ranks second in Southeast Asia in terms of AI startup numbers, after Singapore, with 1,100. Notable Vietnamese AI startups include VinAI, OhmniLabs, Trusting Social, Aimesoft, and Prep.vn. Major corporations like FPT, Viettel, VNPT, and VinGroup have also established their own AI centers, investing in applied research and domestic product development. However, the total investment raised by Vietnamese AI startups remains modest, at only around $80-100 million. There has yet to be a successful Series B funding round, let alone the emergence of an AI unicorn.

One of the critical bottlenecks is human capital. According to the Ministry of Science and Technology (MoST), Vietnam produces about 60,000 IT graduates annually but only around 1,000 have in-depth AI expertise, of which fewer than 300 are considered true AI experts. This vast gap between market demand and talent supply has hindered many AI projects from being effectively implemented.

In terms of computing infrastructure, Vietnam does not yet possess a globally competitive AI supercomputer. Companies must rent GPUs and cloud computing services from foreign providers like AWS, Azure, or Google Cloud, resulting in high costs and dependency risks. Some large firms, such as Viettel, are developing domestic cloud infrastructure, but these remain in pilot stages.

Notably, AI’s diffusion across the economy remains limited. A 2024 survey by the Ministry of Information and Communications (now under MoST) found that only about 12 per cent of small and medium-sized enterprises (SMEs) had experimented with AI applications, and the success rate after one year was less than 5 per cent. The main reasons were a lack of in-house talent, high costs, and insufficient implementation support. This stands in stark contrast to Singapore, where over 50 per cent of SMEs use AI in at least one operational process, thanks to direct support policies (like AI Grants) and public-private advisory networks.

At the local level, only Ho Chi Minh City and south-central Binh Dinh province (now part of Gia Lai province) have plans to develop AI centers. Many cities and provinces have yet to designate AI as a strategic priority, leading to a lack of leadership and scattered resource allocation. This also limits the development of innovation clusters following the “hub-and-spoke” model that countries like South Korea and China have successfully adopted.

In brief, Vietnam’s AI economy is evolving rapidly in terms of speed but the depth of the ecosystem has not yet reached a sustainable level. Unless Vietnam overcomes key constraints, such as talent, infrastructure, capital, and widespread application, the $130 billion target by 2040 will remain an aspiration on paper.

From strategy to legal frameworks

Vietnam was one of the first Southeast Asian countries to issue a National Strategy on Research, Development, and Application of AI, under Decision No. 127/QD-TTg, issued in 2021. The document outlines a vision to 2030, aiming to place Vietnam among the top 50 countries in AI, establish ten innovation centers, and develop at least one high-tech park specializing in AI. However, the gap between strategy and execution remains significant, especially in terms of legal framework, ethical guidelines, and institutional policy-making capacity.

A core issue is the limited institutional space for innovation. New technological experiments, such as large language models (LLMs), AI-as-a-Service offerings, or AI-powered administrative decision-making, still lack a clear regulatory sandbox, unlike in Singapore or the UK. Regulatory bodies remain uncertain about which agencies are responsible for overseeing cross-sectoral AI systems, from healthcare to finance. This lack of clarity discourages many businesses, pushing them to operate “under the radar” rather than openly piloting their innovations.

Another serious bottleneck is the absence of a Law on Personal Data Protection. Currently, Decree No. 13/2023/ND-CP serves only as a temporary measure, lacking effective enforcement mechanisms and an independent oversight body. Meanwhile, personal data is the lifeblood of AI systems, from model training to service personalization. Without strong legislation, the public is unlikely to share data, and businesses cannot ensure accountability. The lack of an entity similar to the European Data Protection Board (EDPB) or the Personal Data Protection Commission (PDPC) in Singapore means that privacy in Vietnam remains a legal grey area.

Furthermore, Vietnam has yet to establish a national AI ethics framework with guiding or binding principles. In contrast, frontrunners like Singapore released their AI Ethics Guidelines as early as 2019, emphasizing fairness, transparency, accountability, and human-centric design. China, despite its centralized governance model, has introduced ethical standards through the Chinese Association for Artificial Intelligence, supported by the government and adopted by enterprises during product development.

In Vietnam, some initiatives are underway. UNESCO is supporting the development of a national AI ethics framework, and the Vietnam Software and IT Services Association (VINASA) has announced the establishment of an AI Ethics Committee. However, these remain advisory in nature, without binding legal force. This poses particular risks as technologies like facial recognition, smart surveillance cameras, and AI-driven content moderation are already being deployed in public sectors without independent oversight, potentially infringing on civil liberties.

Lastly, AI governance still lacks a clear structure. Who is the coordinating authority, MoST or the NIC? The absence of transparent mandates has led to fragmented implementation of the national AI strategy, where many are involved, but no one is held accountable. This is a stark contrast to Singapore, where all AI-related policies are centralized under the National AI Office, which reports directly to the Prime Minister’s Office.

Without a solid institutional framework, Vietnam’s AI economy cannot grow sustainably. Technology infrastructure can be bought, and capital can be borrowed. But trust, the foundation for AI adoption and widespread acceptance, can only be built through responsible policymaking, clear legal frameworks, and consistent ethical standards.

Businesses in AI: Boost or barrier?

In every successful technological development model, from the US to South Korea, from Singapore to Israel, enterprises have always played the central role in bringing technology out of the lab and into real life. In Vietnam, the pressing question is: are businesses truly becoming the driving force of the AI economy, or are they still passive observers?

Vietnam’s AI business ecosystem can currently be divided into three groups: (1) large technology corporations, (2) homegrown AI startups, and (3) companies applying AI as a supplementary tool in operations.

The first group, which includes FPT, Viettel, VNPT, and Vingroup, plays a leading role thanks to their financial resources, human capital, and medium to long-term investment capacity. FPT.AI provides chatbots, virtual call centers, and language processing platforms; Viettel AI develops facial recognition systems, smart cameras, and digital city solutions; and VinAI (a subsidiary of Vingroup) is among the few research and development (R&D) centers in Vietnam that regularly publish internationally-recognized scientific papers.

However, despite their internal strengths, these corporations primarily focus on internal applications or simple packaged products. They have yet to create open, community-driven AI platforms similar to Hugging Face, OpenAI API, or the AI Singapore Foundation. Their products largely remain confined within their own ecosystems, and as a result, they have limited reach to SMEs or the public sector.

Meanwhile, local AI startups like Aimesoft, Trusting Social, OhmniLabs, Elsa, and VUI, though highly innovative, face numerous challenges: difficulties in raising post-Series A funding, a lack of AI-as-a-Service infrastructure, and, especially, the absence of a regulatory sandbox to test their solutions. For instance, a startup aiming to deploy AI in medical diagnostics or credit scoring must navigate long, unclear approval processes, putting them at a disadvantage compared to foreign competitors that can experiment in more permissive regulatory environments elsewhere.

As a result, according to a survey released by the NIC in late 2024, only around 19 per cent of SMEs in Vietnam had attempted to integrate any AI product into their operations, and fewer than 4 per cent maintained such use beyond six months. The main reasons include: (1) high costs, (2) a lack of implementation partners suited to their scale, and (3) skepticism about actual effectiveness.

In contrast, Singapore’s government subsidizes up to 70 per cent of AI pilot costs for SMEs through initiatives like SMEs Go Digital and AI for Industry, while also providing public-private consulting networks under a co-funding model. This approach shortens the go-to-market time for AI products and builds trust among SMEs to adopt advanced technologies.

To sum up, Vietnam’s private sector currently holds both potential and limitations. Without a more proactive role from the government in building institutional frameworks, subsidizing pilot programs, and developing a fair AI application market, the private sector will struggle to become the central engine driving the AI economy.

Golden opportunity

AI is no longer just a trend, it is the infrastructure of a new economic order. Countries leading the AI race are not only gaining technological advantages but also reshaping the global rules of the game, from data standards and machine ethics to the structure of value chains. In this global shift, Vietnam holds unique advantages: a young population, rapid digital transformation, rising tech capacity, and a geopolitically-strategic position that attracts investment.

However, these advantages will quickly erode without decisive strategic action. As history has shown, the “window of opportunity” in any technological wave is short, often lasting only a few years. Without a robust, trustworthy, and self-reliant AI ecosystem, Vietnam risks continued dependency on imported technologies, loss of control over its data, and falling behind in the global tech race.

The $130 billion projection for Vietnam’s AI economy by 2040, as estimated by the NIC, JICA, and BCG, will not materialize on its own. It demands strategic investment, not only in financial terms but in adaptive governance, serious data policies, and a collective commitment to responsible innovation.

An AI-powered economy cannot be built on lofty declarations or ambitious dreams alone. It requires concrete actions: launching regulatory sandboxes, passing a personal data protection law, standardizing AI ethics frameworks, restructuring AI engineering education, and creating an enabling environment for enterprise-led experimentation.

Vietnam now stands at a crossroads: seize AI as a catalyst to leapfrog, boost productivity, and gain control over strategic technologies, thereby securing stable, high growth in the next two decades; or miss the boat and be resigned to a follower’s role - reactive, dependent, and bound by rules set by others.

This is a golden opportunity. But it won’t come twice. Capturing it will require bold action from the government, businesses, and society.