Southeast Asia needs to balance AI innovation and regulation
By Cindy Zheng and Albert J Rapha With the Philippines aiming to propose an artificial intelligence (AI)
regulatory framework during its ASEAN chairmanship in 2026, Southeast Asia is showcasing a proactive
approach to establishing policies and regulations on AI development and deployment.
This step is critical, as a 2020 survey by the consulting firm Kearney found that 70 per cent of Southeast
Asians deemed AI key to their future and urged for its rapid growth in the region.
Devising a comprehensive regional regulatory framework can be a fundamental step in regulating AI
practices and addressing the misuse of this emerging technology. It would bring convergence to ASEAN’s
common ground in developing AI to support the region’s Digital Economy Framework Agreement and the
Post-2025 Agenda.
This may also promote and empower AI’s benefits throughout the ASEAN member states, notably in spurring
competitiveness and fostering innovative regional ecosystems. There is potential to reap up to US$950
billion by 2030, or a 10–18 per cent uplift in regional domestic product from harnessing AI.
The region’s capacity to compete in the global AI market remains hampered by its low level of preparedness
in advancing AI-led technologies. According to Oxford’s Government AI Readiness Index 2023, Singapore is
the most prepared for AI, scoring 81.97 out of 100 — significantly higher than the regional average of
55.49. The discrepancy becomes more apparent when comparing Singapore’s level of preparedness to less
developed countries like Myanmar, which had scoring gaps of 32 per cent and 62 per cent, respectively.
To date, the region’s AI initiatives are also predominantly led by a small number of countries with more
advanced digital economy systems, resulting in a lucid discrepancy in developments related to AI-driven
technologies. Singapore has launched 25 initiatives, including the National AI Strategy, to foster a
trustworthy environment for AI development. This approach extends to the proposal of a Model AI Governance
Framework for Generative AI in January 2024, addressing the rising use of this technology globally.
In addition, Vietnam has launched six initiatives, Thailand has five strategies and Indonesia has one.
Several countries are in the initial stages of developing AI, such as Cambodia, which issued detailed
recommendations on how the government should facilitate and regulate AI in 2023.
While the February 2024 ASEAN Guide on AI Governance and Ethics may serve as a regional best practice
guide for governments and business sectors, the region requires a more granular and comprehensive
framework. This framework should incorporate legal measures to ensure that the development of AI benefits
the region while avoiding its future potential consequences.
Regulating AI is critical to building a safe, inclusive and innovative ecosystem for AI development while
safeguarding ethical values as it is deployed across society. Yet before developing a regional regulatory
framework on AI, ASEAN member states should consider the potential drawbacks of existing regulatory
frameworks, like the European Union’s Artificial Intelligence Act (AIA).
The upcoming ASEAN regulatory framework on AI should avoid the AIA’s dilemma of preventing unlawful
actions in developing and utilising AI while spurring the growth of innovation. Striking a balance between
innovation and regulation is crucial.
Unlike most regulatory approaches, the AIA stands out by utilising a legally binding agreement rather than
relying on voluntary self-regulation of AI developers and companies. While it follows the trend of a
risk-based approach to AI regulation, the European Union — a regulatory power in the digital space since
the General Data Protection Regulation, the Digital Services Act and the Digital Markets Act — is close to
adopting the world’s most comprehensive AI law. This law places strict regulations and requirements on
high-risk and unacceptable categories of AI.
Practitioners fear this might result in over-regulation that could lead to a European brain-drain in the
AI field, as expressed in an open letter signed by over 150 leading European technology firms in June
2023.
The 2023 AppliedAI survey showed that 73 per cent of surveyed venture capitalists from EU member states
expect the act to significantly reduce competitiveness and potentially become a bottleneck in spurring
innovation, notably in the open-source AI industry, which could later dwarf market contestability. This
approach may confine cross-border innovations, resulting in potential backlash due to diverse innovation
approaches in other regions, as well as encountering growing AI nationalism.
In contrast, The European Union continuously stresses that the AIA is based on excellence, supporting
innovation asspecified in the European AI Strategy and the January 2024 launch of the AI innovation
package. A 2023 analysis further estimated that compliance costs for general-purpose AI providers would
likely be minimal, ranging between 0.07–1.34 per cent of total capital expenditure to build AI models.
The AIA could be a regulatory instrument to safeguard European values and ensure a level playing field
with improved resources for companies developing AI systems and tools.
Reflecting on the current double-edged sword of the European Union’s AIA, ASEAN member states should
strive for regulatory convergence among member states with the forthcoming ASEAN AI regulatory framework
and prioritise balancing the innovation and regulation of AI. Following these guardrails, the region may
foster a competitive and inclusive AI ecosystem while upholding ethical values to prevent potential
malicious outcomes from the use of cutting-edge technology.
Albert J Rapha is Research Assistant at German Institute for Global and Area Studies and a Postgraduate
Student at KU Leuven.
Cindy Zheng is a Postgraduate Student at Sciences Po and KU Leuven.