The accelerated integration of Artificial Intelligence (AI) into all manner of devices, services, and platforms has created a palpable sense of urgency for countries to tame and harness AI. This urgency is prompting significant advancements in the field of AI governance, with China, the United States (US), the European Union (EU), and the countries of the G7 taking pivotal steps in shaping the global landscape.In early October, China introduced the Global AI Governance Initiative (GAIGI) for participating countries of the Belt and Road Initiative, emphasising the need for all countries to equally benefit from AI, regardless of their political systems. More recently, US President Biden issued the Executive Order on Trustworthy AI, covering crucial aspects such as safeguards against algorithmic bias and the promotion of innovation in AI.Meanwhile, the EU’s AI Act, the world’s first comprehensive AI law, has entered the last mile of trilogue negotiations, reaching landmark agreements on the provisions concerning the classification of high-risk AI applications. At the same time, the G7 has adopted a voluntary Code of Conduct to mitigate advanced AI systems’ risks and potential misuse, providing guidance to organisations developing generative AI models.As these distinct approaches to AI governance take shape and gain influence, it is important to examine the way other economies – specifically, emerging Southeast Asian digital economies – are defining their own positions.The main question at this time is: Will countries in the region be drawn to a polarised alignment model where they are expected to swear allegiance to one of the dominant approaches; prefer a centralised, supranational model where a multinational, multilateral, and multistakeholder organisation such as the United Nations’ AI advisory Body guides decisions; or lean toward a fragmented, inward-looking model where each country takes its own route based on domestic priorities and interests?Our work with individual Southeast Asian governments and with the ASEAN Secretariat points towards a fourth, much more pragmatic and realist approach; one that simultaneously puts innovative business models at the heart of voluntary risk-based assessments, puts conducive policy and regulatory environments at the top of digitalisation agendas, and puts cross-jurisdictional cooperation and cross-border interoperability at the centre of trade agreement provisions.

The economic promise of AI in Southeast Asia

Our research on AI in Southeast Asia shows there is a thriving and dynamic AI sector in the region.In healthcare, for example, AI-powered systems are increasingly used to improve service provision as well as patient outcomes. In Thailand, the overburdened public healthcare system was able to navigate the worst of the COVID-19 crisis thanks to AI-driven solutions that simplified processes and optimised resource allocation. In Singapore, the National Heart Centre Singapore (NHCS) leverages AI to analyse medical images, aiding in early detection of cardiovascular diseases.AI is also driving innovation and productivity across the region, with AI start-ups changing the way people work and businesses operate. In Indonesia, Advotics helps e-commerce companies overcome the country’s highly fragmented logistics industry by leveraging AI to optimise supply chain operations, improve inventory management, and reduce costs. In Malaysia, Synapse Innovation utilises AI-driven solutions to automate repetitive tasks and streamline business processes – helping a wide range of SMEs boost their overall productivity.AI technologies are also helping overcome traditional language biases in AI development. In Myanmar, local start-up Expa.ai is developing in-house capabilities for e-commerce applications in Burmese, allowing foreign multinationals greater ease of access to Myanmar’s market while creating opportunities for local players to expand into the region.These are only a few examples of thousands of small and large companies putting various types of AI to use across Southeast Asia. It is worth noting, however, that while there are huge domestic technology companies that leverage AI for their digital products and services in the region – Bukalapak, Viettel, Tokopedia, Grab, GoJek, Kata.ai, Shopee – none of these have a dominant regional position comparable to that of OpenAI or Anthropic in the United States.In this context, it is not surprising that AI is considered a key agent of economic growth that must be supported and harnessed through conducive laws, policies, and regulations. As such, most AI plans, strategies, policies, and regulations in the region are prepared with businesses in mind.

Approaching AI as an engine of economic growth

While the processes, approaches, and objectives of AI governance vary from country to country, businesses’ interests – manpower issues, compliance costs, cross-border business models, regional/global expansion, data protection/privacy, cybersecurity – tend to weigh heavily when regulations are put together in Southeast Asia.This is best exemplified by the resolutely business-friendliness of the draft ASEAN Guide on AI Governance and Ethics. While this document is still being reviewed and drafted, it has reportedly been shaped with input from prominent tech corporations such as Meta, IBM, and Google. Among other measures, the Guide is set to include:

• Voluntary guidelines that guide domestic regulations on AI ethics and governance;
• No “unacceptable risk” categories (or similar) that would prohibit certain AI technologies (unlike the EU’s AI Act);
• Advice on the risks of AI (leaving it up to member states to work out the best way to define and address these risks); and
• Guidance for companies to implement an AI risk assessment structure and AI governance training (leaving specific measures up to companies and local regulators).

Business-friendliness is also apparent in the way responsible AI is being operationalised in the region, as governments recognise that in order to deliver on its socio-economic promises, AI must be fully trusted by consumers and understood by businesses.As such, six of 10 ASEAN economies have national AI plans, policies, and strategies in place – the majority of which aim to make the development and usage of AI ethical and responsible. From principles, guidelines, and standards that promote algorithmic transparency, unbiased datasets, and ethical practices to laws, policies, and regulations specifically developed to frame AI adoption, governments across the region are increasingly and explicitly recognising the importance of developing and adopting AI in a fair, ethical, and trustworthy manner.

A melting pot approach to AI governance

Our work with Southeast Asian governments and the ASEAN Secretariat has taught us that lawmakers, policymakers, and regulators in the region are not particularly fond of ‘hegemonic’ approaches coming from bigger players (namely, the EU, the US, and China).From what we know of the draft ASEAN Guide on AI Governance and Ethics, it is said to stray far from the EU’s push for globally harmonised rules that align with its own more stringent measures. This is despite EU officials having toured the region in July 2023 to convince governments to follow the lead of the AI Act, especially on measures such as disclosure of copyrighted and AI-generated content.Regarding China’s Global AI Governance Initiative (GAIGI) for participating countries of the Belt and Road Initiative, it is too soon to evaluate the extent to which this initiative will permeate Southeast Asia’s approach to AI governance. It is possible, however, to surmise that Southeast Asian governments will adhere to its broadest principles that favour AI investment and innovation while securing Chinese funding of projects that modernise ICT infrastructure.When it comes to the US, Singapore’s IMDA and the US’ NIST agencies have recently completed a joint mapping exercise between their respective AI Verify and the AI Risk Management Framework (RMF) programmes. The resulting “crosswalk” document serves three distinct but related objectives:

1. It demonstrates that while they may label or categorise things differently, principles-driven AI governance frameworks can find common ground in terms of approaches and objectives;
2. It illustrates the fact that both Singapore and the US have a resolutely business-friendly stance that is compatible with and conducive for cross-border economic activities; and
3. It positions Singapore and the US as first-mover digital economies that are taking concrete steps towards cross-jurisdictional interoperability, paving the way for other economies to follow suit.

Despite what has been written elsewhere, we do not see this as Singapore fully aligning its approach to AI governance to the US’ own approach; it merely indicates that the US and Singapore both recognise that their approaches are similar enough that they can complement each other and co-exist towards a common goal. Likewise, we do not see this as a sign that the development of AI policy in Singapore (and by extension, in Southeast Asia) is dictated by the expectations of the US; it simply means that policymakers in Southeast Asia and in the US agree that AI is a major and an urgent economic opportunity to seize.This is not to say that Southeast Asia is completely impervious to international initiatives.From what we have seen of the draft ASEAN Guide on AI Governance and Ethics, it seems to indicate a preference for an approach that is broad enough that it allows digital companies and business models to thrive, but structured enough that it respects current and upcoming international commitments.There are indeed various multilateral trade agreements at play, which all require some form of alignment with international approaches. Internationally, there is the Regional Comprehensive Economic Partnership (RCEP) that is already in place, as well as the Indo-Pacific Economic Framework for Prosperity (IPEF) that has just started being negotiated. Regionally, there are ongoing negotiations around the Digital Economy Framework Agreement (DEFA).In addition, several Southeast Asian countries are participating in various international AI agreements; all 10 ASEAN members are signatories of UNESCO’s Recommendation on the Ethics of Artificial Intelligence, which defines ethical AI, outlines key principles, and presents everything that makes for an enabling environment for ethical AI to emerge. Meanwhile, Singapore is an adherent of the OECD/G20 Principles on Artificial Intelligence, and Indonesia (as part of the G20) is a signatory to the G20 AI Principles set out in 2019.Another area that demonstrates indirect influence on the region’s AI governance frameworks is the way that responsible AI frameworks are gradually encompassing complex socio-cultural issues. Our interviews with AI practitioners in the region show that there is a growing socio-cultural push to make AI not just ethical, but inclusive, participative, and accessible. This includes specific consideration for the needs of at-risk and/or marginalised communities, including indigenous populations, ethnic/linguistic minorities, women and children, the elderly, the unconnected, the unbanked, the digitally illiterate, persons with disabilities, those vulnerable to climate change, and those still affected by the economic slowdown caused by the COVID-19 pandemic.Whichever model Southeast Asian governments converge towards, they are unlikely to completely align with a single model or framework. We see them mixing and matching different aspects of the various dominant approaches, choosing the elements that are most conducive and compatible with the economic objectives/interests of their rapidly digitalising economies.Implications for the ASEAN approachAs Southeast Asia navigates the dynamic AI governance landscape, we can expect governments in the region to embrace several key considerations in defining a unified approach:

1. Tailored AI governance approach: SEA governments are unlikely to adhere strictly to a single model or framework. Instead, they are poised to draw from various dominant approaches, selecting elements that align with the rapid digitalisation of their economies and economic objectives. This approach fosters a thriving digital business environment while respecting trade and economic agreements.
2. Business-friendly regulations: The ASEAN Guide on AI Governance and Ethics, shaped in collaboration with tech giants, is oriented towards facilitating business growth. SEA governments are likely to continue adopting a flexible and supportive stance to attract tech investments and foster innovation.
3. Alignment with economic objectives: SEA governments view AI as a significant tool for maximising productivity and efficiency, thereby driving economic growth. As such, they are expected to prioritise AI governance models that harmonise with their broader economic objectives, ensuring policies that promote digital transformation and economic expansion. This approach is vital in a region experiencing rapid digitalisation.

Overall, these considerations will significantly shape the SEA region’s approach to AI governance, as the region seeks to strike a balance between fostering innovation, upholding ethical standards, and ensuring the safety of emerging technologies. In an ever-evolving AI landscape, the ASEAN approach reflects the region’s commitment to realising the full potential of AI while remaining agile in response to shifting global dynamics.For more information on this topic, please contact Jonathan Gonzalez at [email protected] or Li Xing at [email protected].