AI in Southeast Asia

01

Who is using AI

28%
People using generative AIMicrosoft AI Diffusion Report, 2026-Q2
0.5
Claude use per personAnthropic Economic Index, 2026-05
88
ChatGPT messages per person, world rankOpenAI Signals, 2026-Q2

02

What governments are doing

Vietnam

Verified September 29, 2026

Sector rules
  • Draft Circular on safety, risk management and deployment conditions for AI applications in the banking sectorDraft, 2026Circulated for industry comment via Official Dispatch 1162/NHNN-CNTT of 2026-02-13 (6 chapters, 27 articles in the first draft; a later draft reported as 5 chapters, 26 articles), it would require credit institutions, foreign bank branches and payment intermediaries to classify AI systems by risk, keep human oversight, disclose AI interaction to customers, control bias and report serious incidents within 24 hours, with existing systems to comply by 2027-09-01; it was being finalised after Decision 33/2026 as of August 2026 and had not been located as issued on the SBV portal by 2026-09-29.
  • Decision 33/2026/QD-TTg promulgating the List of High-Risk AI SystemsEnacted, 2026Names the specific AI uses in six fields (for example automated credit decisions and transaction controls in banking, surgical robots and diagnostic aids in health, automated student assessment in education, biometric identification in proceedings, autonomous driving and traffic control in transport) that trigger conformity assessment and the other high-risk duties of the AI Law and Decree 142.

Naming: the AI Law's Vietnamese name is Luat Tri tue nhan tao; English sources use 'Law on Artificial Intelligence' or 'AI Law'. Government-portal pages give the law's date as 2025-12-10 (passage and signing), while some English news dated it 2025-12-11. The lead ministry is the Ministry of Science and Technology (MOST, sometimes abbreviated MST on its English site) after absorbing the Ministry of Information and Communications in 2025; the ministry's English site still carries some 'MIC' page paths. A Decree 225/2026/ND-CP (reported in state media as effective 2026-07-01) is said to add AI explicitly to MOST's functions and create a Department of Digital Technology and AI, but this was not confirmed at a primary source. Strategy: the 2021 Decision 127/QD-TTg record on vanban.chinhphu.vn does not yet show a 'replaced' flag, and the government newspaper article on Decision 1671 does not state that it repeals Decision 127, though LuatVietnam reports that it does; treat Decision 1671 as the current strategy and Decision 127 as its predecessor. Political backdrop: Politburo Resolution 57-NQ/TW of 2024-12-22 on breakthroughs in science, technology, innovation and digital transformation (a Party document, not law) drives the whole programme. Law on Digital Technology Industry 71/2025/QH15 (passed 2025-06-14, in force 2026-01-01) originally carried an AI chapter that the AI Law repealed. Pending or unconfirmed: SBV AI banking circular (draft); administrative-sanctions decree for the AI Law and for the PDPL (none located as issued); MOST technical guidance on conformity assessment; the Ministry of Education and Training's AI curriculum framework for grades 1 to 12 from school year 2026-2027 (reported mid-August 2026, no primary document located, so not listed as a sector rule). Penalty amounts under the AI Law are not stated in the sources opened; Vietnam Briefing reports penalties may be set as a share of global revenue but gives no figure. The National AI Ethics Framework is binding only for public-sector AI use and recommended for everyone else; the 2024 Decision 1290 principles were fully voluntary.

03

What researchers are finding

Map of Southeast Asia with Vietnam highlightedVietnam

Jobs and the economy

Applications of Artificial Intelligence in the Vietnamese Economy

Surveys the evidence on how AI is reshaping work, digital skill demand and productivity in Vietnam as the country moves off a low-cost labour model, flagging skills gaps, adoption costs and job quality as the main constraints.

Working paper · July 1, 2026 · IZA Discussion Paper No. 18825

Map of Southeast Asia with Vietnam highlightedVietnam

Government and rules

Viet Nam: Artificial Intelligence Readiness Assessment Report

Assesses Viet Nam's AI ecosystem after the 2025 Digital Technology Industry Law, documenting a 5,000 to 7,000 AI engineer base against a large IT talent shortfall and a fast-growing AI market.

Report · October 1, 2025 · UNESCO AI Readiness Assessment Methodology country report (SHS/BIO/2025/AI-RAM/VN)

Map of Southeast Asia with Southeast Asia, Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand, Timor-Leste, Vietnam highlightedSoutheast Asia, Brunei +10

Government and rules

AI Safety Governance, the Southeast Asian Way

Surveys AI safety and governance moves in every Southeast Asian country, including Timor-Leste, and argues the region is converging on a pragmatic, development-first model distinct from EU or US approaches.

Report · August 25, 2025 · Brookings Governance Studies report

Map of Southeast Asia with Vietnam highlightedVietnam

Language and models

PhoGPT: Generative Pre-training for Vietnamese

PhoGPT-4B is an open Vietnamese generative model and chat variant pretrained from scratch, reported to outperform much larger multilingual models on Vietnamese instruction following.

Preprint · November 6, 2023 · arXiv 2311.02945

04

What people are saying

Map of Southeast Asia with Southeast Asia, Singapore, Indonesia, Thailand, Vietnam highlightedSoutheast Asia, Singapore +3

FulcrumAnalysis

Southeast Asia's Mixed AI Stacks: Pragmatic but Not Risk-Free

Zenobia Chan · September 30, 2026

Shows that regional projects such as SEA-LION, Sahabat-AI, Typhoon and FPT's AI Factory combine American chips, American and Chinese open-weight models and local data rather than picking one bloc. Argues these mixed stacks bring real gains in local relevance and testing capacity, but warns that labels like sovereign can overstate control when every model still runs on the same foreign cloud or chips.

Adapted from her ISEAS Trends in Southeast Asia paper, and written as Washington signals that partners may have to choose between US and Chinese AI coalitions.

Map of Southeast Asia with Southeast Asia, Malaysia, Singapore, Indonesia, Thailand, Vietnam highlightedSoutheast Asia, Malaysia +4

The DiplomatAnalysis

The Global Data Center Boom Has Arrived in Southeast Asia

James Guild · September 29, 2026

Surveys the capital pouring into regional data centres, with Malaysia and the Johor-Singapore zone in front, and asks whether local demand for AI services will ever justify the land, power and water being committed. Frames the boom as a bet whose downside the host countries would carry.

Connects to Malaysian FDI data and the wider debate over whether Southeast Asia is host or beneficiary of the AI build-out.

Map of Southeast Asia with Southeast Asia, Singapore, Vietnam, Malaysia, Indonesia highlightedSoutheast Asia, Singapore +3

ISEAS Perspective (ISEAS-Yusof Ishak Institute)Analysis

Who Can Actually Stop the Machine? The Institutional Test for AI Governance in Southeast Asia

Aditya Vikram Kashyap · September 29, 2026

Argues that the region's growing stack of AI guides, laws and bills says little about whether any institution can actually refuse or halt a deployment, and proposes three tests: a named decision-maker, independence from commercial pressure, and technical capacity to evaluate systems. Compares Singapore's sectoral regulators, Vietnam's new law, Malaysia's proposed bill and Indonesia's pending presidential regulations and finds effectiveness still unproven in all four.

Responds to the wave of 2026 AI rule-making (Vietnam's AI law in force since March, Malaysia's AI Governance Bill, Indonesia's Perpres) by asking who holds the off switch.

Map of Southeast Asia with Southeast Asia, Malaysia, Indonesia, Singapore, Thailand, Vietnam, Philippines highlightedSoutheast Asia, Malaysia +5

The Southeast Asia Desk (Substack)Analysis

The cost of the cloud: Inside Southeast Asia's $30B data centre trilemma

Fredrick Samuel Haryanto and Alvin Qobulsyah · September 28, 2026

Frames the region's AI data-centre rush as a trilemma of power grids, water depletion and community consent, using the Johor protests as the warning case. The explainer argues the build-out will not stop, so the region must decide who bears its costs.

Published as the Jakarta-based newsletter's weekly 'Regional Line' feature.

Map of Southeast Asia with Vietnam highlightedVietnam

Vietnam Ministry of Science and TechnologyOfficial statement

Tie Law Implementation to Real AI Needs: Technology Supports People, People Stay Responsible

Tran Trung Tinh · September 18, 2026

At a Ho Chi Minh City training conference on implementing the AI law, the deputy minister insists AI outputs must be checked before use in public service and that officials, not systems, remain accountable. Directs agencies to start from concrete problems rather than fashionable tools, classify systems by risk, keep records, avoid uploading sensitive data to public platforms, and pilot and share models through provincial science offices.

Operationalises Vietnam's AI law (in force since March 2026) and the ban on uploading classified documents to public AI platforms.

Map of Southeast Asia with Vietnam, Southeast Asia highlightedVietnam, Southeast Asia

Vietnam Ministry of Science and TechnologyOfficial statement

Vietnam Proposes Three Directions for China-ASEAN AI Cooperation

Hoang Minh · September 17, 2026

At the China-ASEAN AI ministerial roundtable in Nanning, Vietnam's deputy minister proposes policy dialogue on AI governance, joint safety testing and risk research with standards suited to local languages, cultures and data sovereignty, and workforce exchange so SMEs and citizens can access AI. Frames all three around a people-centred stance that pairs progress with rights, privacy and accountability.

Vietnam positioning itself, months after its AI law took effect, within China's regional AI outreach.

Map of Southeast Asia with Southeast Asia, Malaysia, Singapore, Vietnam highlightedSoutheast Asia, Malaysia +2

FulcrumAnalysis

For Southeast Asia, the AI Race Is About Access, Not Supremacy

Zenobia Chan · September 16, 2026

Argues that the region's real stake in the US-China AI contest is reliable, affordable, locally usable access rather than frontier capability, and maps dependencies across chips, cloud, power and language coverage. Recommends hedging and diversification to secure what the author calls resilient access.

Adapted from an ISEAS Trends in Southeast Asia paper; sits alongside the Pax Silica vs WAICO pressure on regional governments.

Map of Southeast Asia with Southeast Asia, Malaysia, Singapore, Thailand, Indonesia, Vietnam highlightedSoutheast Asia, Malaysia +4

The DiplomatAnalysis

Will Southeast Asia Actually Benefit From Its AI Boom?

Gerald Mako · September 16, 2026

Notes that Southeast Asia accounts for roughly half of Asia-Pacific data-centre capacity under construction, with a division of labour forming between Singapore and its neighbours, then argues that hosting compute is not the same as building an AI economy. Asks what would turn separate national build-outs into a regional AI market.

Draws on first-half 2026 pipeline figures and the Singapore-Malaysia-Thailand capacity race.

Map of Southeast Asia with Vietnam highlightedVietnam

VnExpress InternationalOpinion

Humans must become wiser to coexist with AI 'species'

Haley Phan · September 16, 2026

After teaching an AI class to 60 Vietnamese students who could list AI's downsides but not how to avoid them, the author argues that treating AI as a tool lowers our guard and that it should be approached as a fast-evolving species. She calls for habits of verification and critical use instead of a brain-off, ChatGPT-on reflex.

Connects to Vietnam's rapid consumer AI adoption (second in ASEAN by Microsoft's count).

Map of Southeast Asia with Vietnam highlightedVietnam

VnExpress InternationalOpinion

The real challenge for university lecturers isn't whether students use AI but how they do

VnExpress reader (university lecturer) · September 5, 2026

A Vietnamese lecturer writes that asking students whether they used AI has become meaningless and that the real gap is faculty preparedness, citing a 2026 study finding nearly nine in ten Vietnamese students use AI tools while only a quarter have any AI-literacy training. The piece calls on universities to redesign courses around guided AI use.

Part of VnExpress' Readers' Views strand on AI in Vietnamese classrooms.

Map of Southeast Asia with Vietnam highlightedVietnam

Vietnam Investment ReviewAnalysis

Vietnam can lead the way in AI regulatory system

Dang The Duc and Le Thi Minh Trang (Indochine Counsel) · September 4, 2026

Two lawyers walk through Decision 33/2026 and Decree 142/2026, which list nearly 50 high-risk AI systems and set self-classification and notification duties under Vietnam's AI Law, the first comprehensive AI statute in Southeast Asia. They argue the framework's risk-based design could make Vietnam a regional reference point if enforcement is proportionate.

Follows the AI Law taking effect in March 2026 and the high-risk list taking effect August 15.

Map of Southeast Asia with Southeast Asia, Vietnam, Singapore, Malaysia, Indonesia, Thailand highlightedSoutheast Asia, Vietnam +4

FulcrumOpinion

Southeast Asia and AI Adoption: The Return of Cold War Strategies

Tobias Burgers and Tran Nguyen · August 31, 2026

Reads Vietnam's AI law and Singapore's local models as Cold War-style hedging between US and Chinese AI ecosystems, then argues that national efforts are too small to matter. Proposes collaborative hedging through a regional public-private consortium modelled on Airbus.

Connects to Malaysia's ILMU model, Vietnam's December 2025 AI Law and the region's sovereign AI turn.

Map of Southeast Asia with Southeast Asia, Vietnam, Indonesia highlightedSoutheast Asia, Vietnam, Indonesia

Tech for Good InstituteOpinion

What Does the World Cup Teach Southeast Asia About Its Digital Future

Tran Thi Thuc Huyen · August 28, 2026

Uses FIFA's equal distribution of AI analytics to all 48 World Cup teams to argue that access to tools does not create capability; people, institutions and locally tailored solutions do. Points to Vietnam's football federation training and Indonesia's community-built Appa Sport platform as models for pairing AI with human judgment.

Part of a run of TFGI pieces arguing that regional AI strategy must invest in skills and institutions, not just infrastructure.

Map of Southeast Asia with Southeast Asia, Vietnam, Malaysia, Singapore highlightedSoutheast Asia, Vietnam +2

The DiplomatOpinion

How Southeast Asia and the Gulf Can Shape the Future of AI

Kelly Forbes, Bruce Li and Hamzah Bin Zaid · August 18, 2026

Argues that two regions that make neither chips nor frontier models are turning their rule-taker status into leverage by pairing Gulf capital with Southeast Asian markets, talent and demand. Points to G42's Vietnam deal and UAE-Malaysia agreements as a template for middle-power AI cooperation.

Follows the February 2026 G42-Vietnam infrastructure framework and the wave of Gulf investment into regional data centres.

Map of Southeast Asia with Vietnam, Southeast Asia highlightedVietnam, Southeast Asia

The DiplomatAnalysis

Vietnam's New AI Law and Its Strategic Implications

Kevin McGahan · July 28, 2026

Assesses Vietnam's state-led AI law, in force since March 2026, as the region's boldest regulatory move and a benchmark against ASEAN's voluntary approach, while arguing Hanoi has ignored the economic, social and environmental fallout of rapid AI transformation. Warns that sovereignty-first regulation can widen inequality.

Responds to implementation of Law 134/2025/QH15 and Vietnam's bid to become a regional AI hub.

Map of Southeast Asia with Southeast Asia, Malaysia, Vietnam, Singapore, Thailand, Indonesia, Philippines highlightedSoutheast Asia, Malaysia +5

Tech for Good InstituteOpinion

The Balancing Act: Harmonising AI Development and Governance in Southeast Asia

Bryan Yeoh Quan Jin and Calvin Woo · July 6, 2026

Argues that with nearly half of regional firms scaling AI beyond pilots, Southeast Asia needs governance that combines ASEAN-level coordination with national rules suited to a multilingual, multicultural region. Surveys the spectrum from Vietnam's binding law to Singapore, Thailand and Indonesia's soft-law approaches and Malaysia's proposed bill.

Written by Malaysia's Centre for the Fourth Industrial Revolution as Malaysia drafts its AI Governance Bill and ASEAN finalises DEFA.