This chapter analyses ASEAN countries’ efforts to make the most out of Artificial Intelligence (AI) for international trade. The results highlight that, while ASEAN Member States have made significant progress through recent AI frameworks, significant disparities and barriers remain affecting their ability to fully take advantage of AI opportunities.
Digital Trade Review of the Association of Southeast Asian Nations
Abstract
Trade and AI are mutually reinforcing. AI is poised to transform the global trade landscape, while open trade and a supportive trade policy are key enablers of AI adoption and diffusion.
ASEAN’s AI readiness is uneven, and trade is critical to access AI technologies. Singapore is a regional leader and global player, but AI technologies are currently concentrated in the USA and China (and, to a lesser extent, the EU).
AMS are making steady progress in adopting and implementing AI initiatives, but at different speeds. Singapore, Viet Nam and Malaysia have established domestic policy frameworks. Indonesia, the Philippines, and Thailand are in advanced discussions. Brunei Darussalam, Cambodia, and Lao PDR are yet to adopt final AI strategies (relying on broader digital economy frameworks covering aspects of AI). In Myanmar, this process is still at an early stage.
Tariffs on “AI goods” remain in Cambodia, Lao PDR, Indonesia, Viet Nam, and Myanmar, affecting the cost of access to critical goods and devices needed for the deployment of AI solutions.
Telecommunications and computer services are particularly restrictive in Lao PDR, Myanmar, Viet Nam (telecommunications), and Cambodia (computer services), affecting access to AI solutions.
Barriers on the movement of AI specialists are also high in Lao PDR, Myanmar, Cambodia and Brunei Darussalam, restricting access to needed AI skills.
Data flow governance remains a bottleneck in Brunei Darussalam, Indonesia and Viet Nam, and lack of data regulation in Cambodia, Myanmar and Lao PDR potentially restricts use and re-use of data powering AI.
However, international co‑operation on issues that relate to AI remains strong due to ASEAN’s strong digital trade provisions in trade agreements.
ASEAN’s ambition for a common regional AI policy framework warrants sustained follow-through. The Digital Economy Framework Agreement provides a key opportunity to address some of these barriers supporting AI adoption and diffusion across the region.
Artificial intelligence (AI) and international trade are interlinked. AI matters for trade because it can increase productivity across all sectors of the economy, enhance supply chain management, and lower trade costs. At the same time, trade and trade policy are critical for the adoption and diffusion of AI systems across economies, facilitating access to AI technologies and the goods, services, people, and data needed to power AI systems (Ferencz, López-González and Oliván García, 2022[100]; WTO, 2025[101]).
ASEAN has recognised that its “ability to capitalise on its strengths in digital and technological transformation” is a critical determinant of future economic success, presenting an opportunity to reinforce the region’s growth and identifying AI as a key driving force (ASEAN, 2025[1]). ASEAN has also understood that the rapid adoption of AI technologies could reinforce barriers to cross-border AI integration if not accompanied by a policy environment that enables AI opportunities to be shared more widely (ASEAN BAC, 2025[2]).
Aiming to address these issues and better take advantage of AI’s opportunities in the region, ASEAN has started to undertake collective action. The ASEAN Guide on AI Governance and Ethics (2024) as well as the ASEAN AI Malaysia Summit and the ASEAN Responsible AI Roadmap (2025) show the region’s growing commitment towards establishing a common policy framework, aiming to adopt evidence-based and coordinated approaches to govern AI and facilitate its adoption, while enabling countries at different stages of development to participate in the adoption and diffusion of AI (Box 5.1).
At the 4th ASEAN Digital Ministers’ meeting, ASEAN released the Guide on AI Governance and Ethics (2024[3]) for responsible AI deployment, emphasising regional interoperability. This voluntary, adaptable framework allows countries to adapt governance to their specific degrees of AI readiness. It was later expanded to include guidelines on generative AI (ASEAN, 2025[4]).
Complementing this framework and developed against the backdrop of the ASEAN Digital Masterplan 2025 and the upcoming DEFA, the ASEAN Responsible AI Roadmap 2025‑2030 (2025[5]) aims to provide ASEAN and national policymakers with actionable steps to implement responsible AI in an integrated and interoperable manner. It offers a step-by-step framework on (1) cross-cutting priorities and foundational policy on the ASEAN level as well as (2) targeted actions which are tailored to the specific circumstances of each Member State. By providing three levels of AI policy maturity to which Member States can assign themselves, it allows for a structured operationalisation of AI in the region.
The ASEAN Committee on Science, Technology, and Innovation (COSTI) is tasked with consolidating AI initiatives, while the Working Group on AI Governance (WG-AI) is responsible for addressing issues such as algorithmic transparency and generative AI governance.
AI development and adoption is projected to contribute between 10% and 18% to AMS’ GDP by 2030 (Kearney, 2020[6]). However, AMS’ position in AI research, development and expertise is not uniform.
AI investment in ASEAN has grown rapidly in recent years. Venture capital (VC) investment in AI and data start-ups in its six largest economies grew from USD 113 million in 2015 to USD 3.8 billion in 2024 (OECD.AI, 2025[7]). However, most investment is concentrated in regional hubs, namely Singapore (USD 12.7 billion, 80% of all ASEAN AI venture funding from 2015 to 2024), and to a lesser extent Indonesia (USD 1.6 billion) (Figure 5.1).
Similar patterns are present in the digital infrastructure needed for the development and deployment of AI solutions. Investments in AI-ready data centres have exceeded USD 50 billion in ASEAN, especially in Malaysia (USD 25 billion), Singapore (USD 9 billion), and Thailand (USD 8 billion) ((Ajmone Marsan et al. (2024[8]); Google, Temasek and Bain & Company (2024[9])).
Granted AI patents per 100 000 inhabitants show that Singapore was the best-performing ASEAN economy in 2023 (ranked 7th worldwide), followed by Malaysia (34th), Viet Nam (49th), Thailand (51st), the Philippines (53rd), and Indonesia (55th) (Stanford University, 2025[10]). Singapore’s AI publications per capita are three times as high as the second-best performing AMS, Brunei Darussalam (Figure 5.2, Panel C). Where skill levels are concerned, Singapore has 3.5 AI professionals per 1 000 workers, compared to Malaysia’s 0.5, and roughly 0.2 for the Philippines, Thailand, Indonesia, and Viet Nam (Meyer et al., 2025[11]).
However, despite the relatively strong position of some AMS, their role remains modest when compared to economies leading in AI investment, innovation, and skills. AI VC investment is highly concentrated in the United States, which received a sum of USD 594 billion from 2015 to 2024, significantly surpassing China (USD 275 billion), the EU27 (USD 75 billion), and all other economies (OECD.AI, 2025[7]).
The origin of AI inventions, too, is concentrated in a few countries. China, the United States, Japan, and the EU27 contributed to almost 80% of AI patents in 2020‑23 (OECD, 2025[12]) (Figure 5.2, Panel A). Similarly, when it comes to AI publications, most of these are concentrated in China, the United States, and the EU27 (Figure 5.2, Panel B). Lastly, as concerns the demographics of AI skills, the majority of data scientists and professionals in machine learning is located in Europe and North America (OECD.AI, 2025[13]).
This concentration of knowledge, skill and investment underscores the need for open markets to enable a greater diffusion and adoption of AI technologies.
Venture capital (VC) investments in AI by country in USD million
Source: OECD.AI (2025), data from Preqin, last updated 01 October 2025, accessed on 07 October 2025, https://oecd.ai/.
Note: Data for Panel A refer to patent applications filed under the Patent Cooperation Treaty (PCT), by priority date and inventor’s location, using fractional counts. Technology domains are defined by the OECD. Data for Panel B and C are averaged over 2022‑24.
Source: OECD, STI Micro-data Lab: Intellectual Property Database, http://oe.cd/ipstats, December 2025 (Panel A); OECD.AI (2025), data from OpenAlex, last updated 30 September 2025, accessed on 22 December 2025, https://oecd.ai/ (Panel B and C).
Given the importance of AI in enhancing ASEAN’s future economic success and considering the crucial role trade plays in that regard, this section evaluates how AMS’ trade policies can be leveraged to increase the adoption and diffusion of AI. The aim is to help AMS make the most out of AI’s opportunities while also addressing challenges raised by the emerging technology. This is done by looking across the different areas that matter for AI and trade, starting with domestic policies, then measures affecting access to the hardware and devices that power AI use, followed by considerations related to access to services, skills and data flows, as well as AMS’ track record in addressing AI as part of trade agreements.
5.3.1. Domestic AI policy frameworks vary considerably across AMS
While ASEAN regional initiatives demonstrate a shared commitment to developing and implementing AI frameworks, approaches and priorities vary considerably across AMS (Figure 5.3). To date, six out of ten AMS have adopted explicit “AI strategies”.1 The remaining four, although in the process of developing more targeted discussions on AI, currently rely on broader digital economy strategies with limited AI focus:
Malaysia, Singapore, and Viet Nam have the most developed domestic policy frameworks governing AI, having established comprehensive and transparent guidelines and/or rules for its development and deployment.
Indonesia, the Philippines, and Thailand have set out a detailed set of policies to address AI opportunities and challenges, guided by comprehensive AI strategies, however they are at earlier stages of implementation of these strategies.
Brunei Darussalam, Cambodia, and Lao PDR are yet to adopt a final AI strategy, currently relying on their broader digital economy frameworks instead, but respective discussions are ongoing. In Myanmar, this process is still in an early stage.
In terms of domestic policy frameworks governing AI development and deployment, four stages of development can be distinguished.
Beginning discussions refers to countries which do not explicitly mention AI in their digital economy frameworks. This is part of a more general context in which discussions about developing an AI-specific strategy to guide policy initiatives are still in early stages.
Ongoing discussions includes countries that highlight AI within broader digital economy frameworks. These countries do not have a finalised AI strategy yet but have started discussions on one.
Advanced discussions encompasses countries which have adopted an explicit AI strategy, serving as a blueprint for developing and deploying AI. It includes comprehensive policies to address identified AI opportunities and challenges, including ethics, talent, computing and data, and the broader AI ecosystem. Countries in this category have not started, or are in earlier stages of, implementing this blueprint or action plan.
Established framework relates to those countries that – building upon their AI strategy – have established a comprehensive and predictable AI governance framework, which may take the form of voluntary frameworks, guidelines, or AI-specific binding laws. Countries in this category are thus implementing the strategy.
As of December 2025
Source: based on AI policy initiatives evaluated applying the classification criteria outlined in Box 5.2.
Beginning discussions
Myanmar does not have a comprehensive AI policy framework in place yet, although the country’s Digital Economy Roadmap (Digital Economy Development Committee, 2018[14]) mentions relevant elements, including with respect to open data and data sharing but is largely focused on promoting basic digital infrastructure. Myanmar’s Cybersecurity Law No. 1/2025, which impacts digital platforms and data governance might also have bearing on AI applications. Plans are, however, underway to draft a National AI Strategy, with a focus on ethical guidelines, developing human resources in AI, and fostering international collaboration.2 To this end, high-level coordination meetings are currently being held.
Ongoing discussions
Brunei Darussalam currently addresses AI within its Digital Economy Masterplan 2025 (Digital Economy Council, 2020[15]), seeking to transform the country into a ‘Smart Nation’. Brunei recently launched its Guide on AI Governance and Ethics (Authority for Info-communications Technology Industry (AITI), 2024[16]), which sets out national principles for AI adoption aligned with ASEAN discussions and relying on voluntary guidelines. AI is set to be central to Brunei’s next Digital Economy Masterplan prioritising infrastructure development, talent, and innovation, with a national AI strategy being prepared to coordinate efforts (Rasidah, 2025[17]).
Cambodia’s approach is focused on early-stage integration of AI into its national development agenda, prioritising setting up foundational digital infrastructure as a prerequisite for comprehensive AI adoption through the Digital Economy and Society Policy Framework 2021‑2035 (Government of Cambodia, 2021[18]). In June 2025, consultations were launched on a draft National AI Strategy 2025‑2030, outlining flagship activities with particular emphasis on infrastructure and human resource development, public sector adoption, and standardisation. In its draft, Cambodia recognises it is in the early stages of AI development and deployment, including limited access to data, computing resources, and talent (Ministry of Post and Telecommunications, 2025[19]).
In Lao PDR, AI is highlighted as an important element in the National Digital Economy Development Strategy (2021‑2030) and the related five-year Plan (2021‑2025).3 As part of these, Lao PDR has launched several foundational initiatives related to AI, including a project in co-operation with China on developing Lao PDR’s first large-scale AI system (Vientiane Times, 2025[20]), and a feasibility study on the creation of an AI Special Economic Zone (World Internet Conference, 2025[21]). The Ministry of Telecommunications and Communications is leading the process of drafting a formal AI strategy aiming to promote technological innovation while considering ethical concerns (Asia News Network, 2025[22]).
Advanced discussions
Indonesia’s National AI Strategy 2020‑2045 (Agency for the Assessment and Application of Technology (BPPT), 2020[23]) includes policies aimed at establishing data sharing practices, creating implementation guidelines for the ethical use of AI, strengthening the implementation of established rules (area 1), training and recruiting AI talents (area 2), providing large-scale computing power and shared data resources (area 3), as well as developing a national AI ecosystem of collaborative research and innovation (area 4). Building on the 2020 strategy, the 2025 National AI Roadmap (Ministry of Communications and Digital, 2025[24]) is the country’s blueprint for adopting and developing AI technology, accounting for recent developments in generative AI. It incorporates three main action plans: AI ecosystem development (national talent development; research, including an open sandbox platform; expanding digital infrastructure, including high-performance computing and a national cloud hosted in “sovereign data centers”), AI development in priority sectors, and AI financing, with an initial focus on, among others, data and computing.
The Philippines launched the National AI Strategy Roadmap (NAISR) 2.0 in 2024 (Department of Trade and Industry, 2024[25]), building on the National AI Roadmap of 2021. It identifies limited computational and human resources, a lack of capacity to develop data strategies, as well as legal uncertainties as barriers to wider AI adoption. It proposes improving availability of fast internet connectivity and of data, nurturing current and future AI talent, and accelerating innovation, including by attracting investment and foreign technology professionals, ensuring a level playing field, simplifying business regulations and setting clear guidelines, and consistent IP governance. The high-level 2025 National AI Strategy focuses on similar key areas and includes plans for a 26-fold increase in high-performance computing (HPC) power by 2028 (OpenGov Asia, 2025[26]).4
Thailand launched its National AI Strategy and Action Plan in 2022 (Ministry of Digital Economy and Society, 2021[27]) to enhance R&D infrastructure (national data platforms, high performance computing, open-source AI platforms), human capability, and legal preparedness by 2027. The government is prioritising the development of AI expertise required for an effective AI ecosystem. Thailand has relied on voluntary standards, such as an AI Ethics Guideline. In 2024, the Ministry of Digital Economy issued guidelines on generative AI. The Draft Principles of the AI Law, taking a risk-based approach to AI governance, are currently under consideration (Norton Rose Fulbright, 2025[28]). Inter alia, it provides for AI regulatory sandboxes and an AI Governance Center.
Established framework
Malaysia’s AI efforts are anchored in the National AI Roadmap 2021‑2025 (Ministry of Science, Technology and Innovation, 2022[29]) and its forthcoming successor, the National AI Action Plan 2030, which is currently under discussion. It aims to strengthen digital infrastructure (cloud, data sharing, network connectivity), attract talent, and advance R&D. Malaysia has adopted the National Guidelines on AI Governance and Ethics (AIGE), which propose voluntary ethical standards for AI use and are intended to be aligned with international principles, including the OECD AI Principles. Malaysia’s National AI Office (NAIO), established in 2024, serves as a central coordinating body for AI policy, while ensuring alignment with international standards.
Singapore has developed a comprehensive, “agile” AI framework based on its National AI Strategy (NAIS, published in 2019 and updated in 2023 as NAIS 2.0) (Government of Singapore (2019[30]) (2023[31])). It outlines the aim to build an AI ecosystem to drive innovation and growth, relying on voluntary frameworks and guidelines for the responsible development and use of AI, both sector-specific and horizontal (Model AI Governance Framework, Model AI Governance Framework for Generative AI). NAIS 2.0 has the aim of both developing specialised areas of excellence in major economic sectors to generate value (expertise), as well as enabling people and organisations to use AI in a responsible way (empowerment). These initiatives also include significant investments in AI, such as a government commitment to spend over USD 1 billion to develop its AI ecosystem over five years. Singapore has also established its own AI Safety Institute, with a broad mandate that includes technical testing, research, policy development, and international co‑operation, as well as the AI Verify Foundation, which aims to encourage responsible AI practices by companies.
Viet Nam’s National Strategy for AI Research, Development, and Application through 2030 (Prime Minister of Viet Nam, 2021[32]) emphasises AI’s role in enhancing competitiveness. It develops strategic directions and targets: building a coherent policy framework for AI, promoting data sharing, establishing open databases, establishing a national centre for high performance computing and cloud infrastructure, human resource development, attracting investment, and promoting adaptation. The Law on Digital Technology Industry (Viet Nam, 2025[33]) took a risk-based approach to AI, aiming to promote innovation and safeguard national security and data integrity. In late 2025, Viet Nam consolidated its governance of AI development and deployment under a single framework by adopting its Law on AI (Ministry of Science and Technology, 2025[34]), replacing most AI-related provisions of the Digital Technology Industry Law. It is inspired by the EU AI Act and outlines a risk-based classification system, but places stronger emphasis on national sovereignty considerations, data autonomy, and infrastructure (Vietnam Briefing, 2025[35]). It also tasks the Prime Minister with reviewing Viet Nam’s AI Strategy at least every three years or upon significant market or technological developments. A National AI Development Fund to support developing AI capabilities, and a sandbox mechanism are also foreseen.5
5.3.2. Tariff-barriers on AI enabling goods remain high in some AMS
Cost-efficient access to AI enabling goods is essential for economies seeking to deploy AI systems. Tariffs are likely to affect access to and the cost of products critical to AI applications. These products encompass a broad spectrum, ranging from raw materials for AI hardware (e.g., palladium) to processed chemicals like silicon used in semiconductors, which serve as intermediate inputs for AI systems’ data storage, as well as equipment for automatic data processing machines. A non-exhaustive list of AI enabling products can be found in the World Trade Report by the WTO (2025[36]).
The average ASEAN ad-valorem tariff on AI enabling goods is around 3.6%, which is notably lower than the global average of 5.3% (Figure 5.5). Singapore and Brunei Darussalam apply tariffs close to zero, while Cambodia stands out with an average rate exceeding 10% – the highest among all AMS analysed for the year 2022. Keeping tariffs low will be critical in making the most out of the AI revolution.
Ad-valorem equivalent tariff (AVE) in %, 2022
Note: This graph shows applied ad-valorem equivalent (AVE) tariffs on AI enabling goods in 2022 in HS nomenclature 2022. Tariffs reflect averages across partner countries and AI enabling goods. The list of these goods is taken from the WTO World Trade Report (2025[36]).
Source: MacMap 2022 (Guimbard et al., 2012[37]).6
5.3.3. Services enable access to AI technologies
Telecommunications and computer services are the backbone of well-functioning AI systems. Indeed, AI relies on critical telecommunications infrastructure, which ensures efficient and reliable internet connectivity for the transmission of data, sound, and video. AI systems also depend on computer services, whether to access AI technologies directly, but also for the design and development of software and computer systems, as well as for data processing and database management.
However, as shown in Chapter 2, telecommunications and computer services continue to face significant trade barriers in AMS. While barriers in the telecommunications sector primarily take the form of competition-related impediments and restrictions on foreign market entry, barriers in computer services are largely characterised by restrictions on the movement of people and on foreign market entry.
Despite regulatory reforms in some AMS leading to a reduction in barriers to the provision of telecommunications and computer services over the last decade, barriers remain relatively high in the ASEAN region compared with the OECD average. In 2024, ASEAN recorded an average STRI score in telecommunications of almost three times the OECD average, and almost 50% higher in computer services.7
Telecommunications services generally face a higher regulatory restrictiveness in AMS than computer services. In both sectors, Singapore, Malaysia, and the Philippines rank among the least restrictive AMS, while Lao PDR and Myanmar remain among the top three most restrictive ones, and Brunei Darussalam, Thailand, and Indonesia, fall within the middle range. Viet Nam has the highest level of restrictiveness in telecommunications but ranks in the mid-range for computer services among AMS, while Cambodia is among the three most restrictive countries in computer services.
5.3.4. Skilled workers
Building robust AI ecosystems relies on technical and interdisciplinary skills that can be complicated to source locally given gaps in AI knowledge and skills. While digital transformation is advancing across ASEAN, it exposes existing talent shortages (Mahusin and Prilliadi, 2025[38]), as enumerated in existing AMS AI strategies. Although opportunities for remote work have increased in recent years, most advanced AI roles still require cross-border movement, which remain affected by barriers such as the need to demonstrate the lack of suitable local labour supplies, burdensome application procedures, and short permitted stays (Ferencz, López-González and Oliván García, 2022[39]).
Talent mobility is a key tenet of ASEAN’s efforts towards regional economic integration. For instance, the AEC Strategic Plan 2026‑2030 emphasises facilitating and advancing digital talent mobility as one of its strategic measures (ASEAN, 2025[1]). In recent years, AMS have developed several important initiatives to facilitate cross-border mobility of people for the provision of services. Within ASEAN, the ASEAN Agreement on the Movement of Natural Persons (MNP Agreement, signed in 2012) provides a framework for the successive elimination of restrictions on the temporary cross-border movement of AMS nationals as business visitors, intra-corporate transferees, or contractual service suppliers (“Mode 4”). It supersedes previous Mode 4 commitments made under the ASEAN Framework Agreement on Services (AFAS). However, commitments vary by country and practical implementation faces challenges, such as visa standards that continue to be not uniform (Tan, Detros and Claudio, 2024[40]).
The MNP Agreement also includes mechanisms for the recognition of professional qualifications. The ASEAN Trade in Services Agreement enhances the existing AFAS but is superseded by the MNP Agreement. Furthermore, ASEAN has developed the ASEAN Qualifications Reference Framework (AQRF) which helps compare education and training across AMS. Beyond the ASEAN level, many AMS have developed national visa programmes to attract high-skilled tech talent, such as Indonesia, Malaysia, Thailand, and Singapore, and ideas for region-wide ‘tech passes’ aimed at digital professionals have been brought forward.
Based on the STRI scores for the movement of people in computer services and telecommunication, most AMS have a more restrictive framework for the movement of AI professionals than the OECD average (Figure 5.5). Lao PDR, Myanmar and Cambodia, and to a lesser extent Brunei Darussalam, have the most restrictive environment for AI-relevant talent mobility, all well above the ASEAN average. In recent years, domestic restrictiveness has remained relatively stable, with the Philippines as a notable exception. In general, restrictiveness to talent mobility in computer services is considerably higher than in telecommunication services.
STRI sub-scores for the movement of people, in 2014 and 2024
Note: STRI scores for Brunei Darussalam, Cambodia, Lao PDR, and Myanmar refer to the year 2023 (latest available year) instead of 2024. For these countries, ASEAN average is based on 2023 instead of 2024.
Source: OECD Services Trade Restrictiveness Index (STRI) (2024).
5.3.5. Data governance
Data is one of the main ingredients of AI systems, but the emerging patchwork of regulatory approaches to data flows risks reducing the ability to deploy AI systems globally by reducing access to data and requiring duplication of activities. More interoperable regulatory approaches that enable the free flow of data with trust are needed to help reap the benefits of AI.
With training AI models requiring access to large datasets, more data is crossing international borders. Furthermore, the deployment and improvement of AI requires continuous exchange of data to make real-time interconnectedness possible, which often needs to cross international borders. Therefore, fostering free cross-border data flows with trust becomes crucial for countries aiming to leverage the benefits of AI.
However, as is shown in Chapter 3, the regulatory landscape that currently underpins cross-border data flows in ASEAN is complex and for some AMS becoming increasingly restrictive. At present, the Philippines, Singapore, Malaysia and Thailand are the only AMS with domestic regulation aligned with the concept of Data Free Flows with Trust (DFFT). Brunei Darussalam, Indonesia and Viet Nam have adopted more restrictive, ad-hoc, approaches to cross-border data flows which are likely to inhibit the ability to fully utilise and adopt AI solutions. In turn, Cambodia, Myanmar and Lao, have yet to adopt regulation in this area which can raise concerns about how data is safeguarded, and the conditions needed to use and transfer data critical for training AI models.
Although there is progress in terms of cross-border data flow provisions in trade agreements and in inter-governmental arrangements such as the ASEAN Model Contract Clauses and the Global CBPR, more work needs to be done in the region to create an enabling environment that promotes the use and re-use of data, including and especially for developing AI solutions that can help AMS meet future opportunities.
5.3.6. International co-operation
As AI technologies continue to evolve, there is a need for concerted international efforts to update governance frameworks to new realities. At present, regional trade agreements (RTAs) already address key AI-relevant issues, including cross-border data flows, privacy and data protection, location of computing facilities, cybersecurity, open government data, and protection of source code. More recent RTAs and Digital Economy Agreements (DEAs) also incorporate provisions explicitly addressing AI, which recognise the importance of developing frameworks for the trusted, safe and responsible use of AI (Ferencz, López-González and Oliván García, 2022[39]). In this respect, given the strong openness of the region, performance in these areas remains strong across most AMS (see Chapter 4).
Signed as of September 2025
Note: Item captures AI as a distinct provision or as part of a broader co‑operation/emerging technologies provision. ‘Hard’ commitments are those that a party is obliged to comply with, whereas ‘soft’ commitments are those that cannot be enforced by another party.
Source: based on the TAPED database version of September 2025 (Burri and Vásquez Callo-Müller, 2025[41]).
Currently, Singapore is the only AMS that has signed RTAs that include specific AI provisions (five in total, co-leading the countries with the most signed RTAs that contain AI provisions). Although these provisions are generally ‘soft’ in nature, they aim to promote international dialogue and co‑operation on this frontier issue, without making more specific, binding commitments.
For many AMS, the Digital Economy Framework Agreement (DEFA) will be a first opportunity to discuss AI in the context of trade discussions, potentially improving policy coordination across the region and promoting open digital markets. According to ASEAN (2025[42]), broad agreement has been reached on emerging technologies (including AI), although the DEFA remains to be finalised by 2026.
Many of the challenges identified in the context of digital trade are today reinforced in the case of AI. Tariffs, services regulations, as well as data governance frameworks have the potential to profoundly impact the availability and diffusion of AI and enabling goods and services. These constrain the capacity of economies to harness the potential of AI (WTO, 2025[36]).
Within ASEAN, which includes countries at various levels of AI policy elaboration and adoption, there are a number of actions that could help put in place a trade policy environment conducive to the wider adoption and diffusion of AI. Efforts to ease relevant trade barriers, create an enabling environment for trade in services, align rules on data flows, and facilitate talent mobility could have a meaningful impact in narrowing the AI divide within ASEAN and advance its positions in the global AI environment. An ambitious DEFA could play an important role to ensure more interoperable and open policies.
As of December 2025
Note: Table excludes initiatives that have not been adopted yet, such as draft AI strategies and laws.
References
[23] Agency for the Assessment and Application of Technology (BPPT) (2020), National Strategy for Artificial Intelligence 2020 - 2045.
[8] Ajmone Marsan, G. et al. (2024), Towards an ASEAN Innovation Ecosystem: Start-up Creation for Inclusive and Sustainable Economic Development.
[42] ASEAN (2025), “ASEAN Economic Community Council Statement on The Substantial Conclusion of the ASEAN DEFA Negotiations”, https://asean.org/wp-content/uploads/2025/10/ADOPTED-AECC-Statement-on-Substantial-Conclusion-of-DEFA-Negotiations-24Oct2025.docx.pdf (accessed on 18 December 2025).
[1] ASEAN (2025), ASEAN Economic Community Strategic Plan 2026-2030.
[5] ASEAN (2025), ASEAN Responsible AI Roadmap 2025-2030.
[4] ASEAN (2025), “Expanded ASEAN Guide on AI Governance and Ethics - Generative AI”.
[3] ASEAN (2024), ASEAN Guide on AI Governance and Ethics.
[2] ASEAN BAC (2025), Artificial Intelligence (AI) and Digital Transformation in the ASEAN Region, https://asean-bac.org/news-and-press-releases/artificial-intelligence-%28ai%29-and-digital-transformation-in-the-asean-region.
[22] Asia News Network (2025), AI seen to advance digital governance, industry, ethics in Laos, https://asianews.network/ai-seen-to-advance-digital-governance-industry-ethics-in-laos/.
[16] Authority for Info-communications Technology Industry (AITI) (2024), Guide on Artificial Intelligence (AI) Governance and Ethics for Brunei Darussalam.
[41] Burri, M. and M. Vásquez Callo-Müller (2025), TAPED: Trade Agreement Provisions on Electronic Commerce and Data, http://ptas.mcgill.ca/;.
[25] Department of Trade and Industry (2024), National AI Strategy Roadmap 2.0 (NAISR 2.0), https://erikalegara.com/uploads/NAISR2.0_July2024.pdf (accessed on 19 December 2025).
[15] Digital Economy Council (2020), Digital Economy Masterplan 2025, https://www.mtic.gov.bn/DE2025/documents/Digital%20Economy%20Masterplan%202025.pdf (accessed on 18 December 2025).
[14] Digital Economy Development Committee (2018), Myanmar Digital Economy Roadmap.
[39] Ferencz, J., J. López-González and I. Oliván García (2022), Artificial Intelligence and International Trade: Some Preliminary Implications, OECD.
[9] Google, Temasek and Bain & Company (2024), e-Conomy SEA 2024.
[18] Government of Cambodia (2021), Cambodia Digital Economy And Society Policy Framework 2021 - 2035, https://asset.cambodia.gov.kh/mptc/media/EN-Policy-Framework-of-Digital-Economy-and-Society.pdf (accessed on 18 December 2025).
[31] Government of Singapore (2023), Singapore National AI Strategy 2.0, https://file.go.gov.sg/nais2023.pdf (accessed on 19 December 2025).
[30] Government of Singapore (2019), National Artificial Intelligence Strategy, https://file.go.gov.sg/nais2019.pdf (accessed on 19 December 2025).
[37] Guimbard, H. et al. (2012), MACMAP-HS6 2007, AN EXHAUSTIVE AND CONSISTENT MEASURE OF APPLIED PROTECTION IN 2007.
[6] Kearney (2020), “Racing toward the future: artificial intelligence in Southeast Asia”.
[38] Mahusin, M. and H. Prilliadi (2025), “Enhancing ASEAN’s Competitiveness: Strategies for Talent Mobility and Cooperation in the Digital Economy”, https://asean.org/scholarship/.
[11] Meyer, M. et al. (2025), Unlocking Southeast Asia’s AI Potential, BCG.
[24] Ministry of Communications and Digital (2025), National Artificial Intelligence Roadmap.
[27] Ministry of Digital Economy and Society (2021), National AI Strategy and Action Plan (2022 – 2027).
[19] Ministry of Post and Telecommunications (2025), Draft National Artificial Intelligence Strategy 2025-2030, https://go.gov.kh/mptc/1w17yx (accessed on 18 December 2025).
[34] Ministry of Science and Technology (2025), First-ever Law on Artificial Intelligence approved, https://english.mst.gov.vn/first-ever-law-on-artificial-intelligence-approved-197251215231241888.htm.
[29] Ministry of Science, Technology and Innovation (2022), Malaysia National Artificial Intelligence Roadmap 2021-2025.
[43] Ministry of Technology and Communications (2021), National Digital Economy Development Vision, Strategy, and Plan, https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwj71I_wpseRAxXdnycCHZuuGoEQFnoECBgQAQ&url=https%3A%2F%2Fmtc.gov.la%2Findex.php%3Fr%3Dsite%2Fdownloadfile%26file%3DWi_AjMvVbw1a5dfyK2hewd%2Fd8041fafe380173c81841cdce0be35b4.pdf&usg=AOvVaw1fakfrcrwzZMUvcFXLVWxP&cshid=1766066694153492&opi=89978449 (accessed on 18 December 2025).
[28] Norton Rose Fulbright (2025), Thailand’s draft AI law: A new era for governance and innovation, https://www.nortonrosefulbright.com/en/knowledge/publications/0fec663b/thailands-draft-ai-law-a-new-era-for-governance-and-innovation.
[12] OECD (2025), Identifying Emerging AI Technologies Using Patent Data, https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/09/identifying-emerging-ai-technologies-using-patent-data_5c8da861/d17e9a1a-en.pdf (accessed on 23 December 2025).
[13] OECD.AI (2025), Developers in Stack Overflow survey by country, https://oecd.ai/en/data?selectedArea=ai-demographics.
[7] OECD.AI (2025), VC investments in AI by country, https://oecd.ai/en/data?selectedArea=investments-in-ai-and-data&selectedVisualization=vc-investments-in-ai-by-country.
[26] OpenGov Asia (2025), The Philippines: President Approves Comprehensive AI Roadmap, https://archive.opengovasia.com/2025/05/22/the-philippines-president-approves-comprehensive-ai-roadmap/.
[32] Prime Minister of Viet Nam (2021), National Strategy On R&D and Application of Artificial Intelligence.
[17] Rasidah, H. (2025), AI will be central to next digital master plan, The Scoop, https://thescoop.co/2025/06/24/ai-will-be-central-to-next-digital-master-plan/ (accessed on 18 December 2025).
[10] Stanford University (2025), Artificial Intelligence Index Report 2025.
[40] Tan, M., K. Detros and B. Claudio (2024), “Opportunities for Talent Collaboration in ASEAN in the Digital Age”, The ASEAN 41/2024, http://www.asean.org.
[20] Vientiane Times (2025), Laos launches pioneering national AI system to fuel digital economy, https://www.vientianetimes.org.la/freefreenews/freecontent_200_LaosLaun_y25.php.
[33] Viet Nam (2025), Law on Digital Technology Industry.
[35] Vietnam Briefing (2025), “Formalizing Artificial Intelligence Governance: Vietnam’s First AI Law”, https://www.vietnam-briefing.com/news/vietnams-ai-law-regulatory-milestone-business-implications.html/.
[21] World Internet Conference (2025), Laos to develop AI infrastructure with green energy, https://www.wicinternet.org/2025-06/06/c_1098704.htm.
[36] WTO (2025), World Trade Report 2025: Making trade and AI work together to the benefit of all, http://www.wto.org.
← 1. “Explicit AI strategies” refer to established domestic strategies, blueprints or frameworks that set out goals and mechanisms to deal with AI.
← 3. These are anchored in a twenty-year Vision (2021‑40) guiding digital transformation initiatives. See Ministry of Technology and Communications (2021[43]).
← 4. While the Philippines is still in the process of formulating a comprehensive and transparent AI governance framework, it has introduced targeted measures to address issues emerging from AI. Proposed bills include the creation of a national authority to lead national AI efforts and measures to address labour displacement and data privacy.
← 5. Foreign suppliers of high-risk AI systems are required to have a local contact point in Viet Nam. If the system is subject to mandatory conformity verification, the supplier must additionally establish a commercial presence or appoint an authorised representation.
← 6. Market Access Map, International Trade Centre, www.macmap.org.
← 7. In telecommunications, ASEAN scored an average STRI of 0.45 out of 1 (down from 0.51 in 2014) compared with an OECD average of 0.17, which has marginally increased since 2014. In computer services, ASEAN’s STRI amounted to 0.26 (down from 0.28 in 2014) compared with an OECD average of 0.18 (up from 0.17 in 2014).