INTRODUCTION

The ASEAN Guide on Artificial Intelligence (AI) Governance and Ethics (ASEAN AI Guide) was endorsed at the Fourth ASEAN Digital Ministers Meeting (ADGMIN),which concluded on 2 February 2024. Premised on the importance of nurturing a “trusted AI ecosystem”, this Guide is based on seven guiding principles including transparency and simplicity, fairness and equity, security and safety, alongside other salient aspects. The discussion on the need for suitable governance levers, or at the very least, for closer tabs on AI technologies, greatly accelerated with the exponential growth of generative AI applications such as ChatGPT in 2023. Though these developments represent exciting technological breakthroughs, they also raise concerns over ethical and governance risks. Moreover, recognizing these risks and how to best mitigate them, is a matter of some urgency, given their rapid proliferation.

Some jurisdictions are attempting to implement heavy-handed legislation such as the European Union (EU) AI Act, but compared to these, the ASEAN AI Guide is a notably more light-touch approach to managing the AI space that is more strategically apt for the region. With the absence of a central legislative body in ASEAN, and with the diversity in digital capabilities and regulatory capacities among ASEAN member states (AMS), a more ambitious endeavour would miss the mark. Moreover, it has been emphasized at its release that business innovation is being prioritized over other factors and thus a less heavy-handed approach would be a better fit.

The ASEAN AI Guide forms a part of the ASEAN Digital Masterplan 2025 (ADM2025) which is grounded in the establishment of “secure and transformative digital services, technologies and ecosystem”. While the Guide may seek to align with international best practices and standards, it is meant to serve more as a “practical guide” for organizations who bear the onus over the responsible and ethical development and deployment of AI systems, rather than as a policy playbook for governments. As such, governments and regional bodies feature as secondary facilitative actors, with the Guide homing in on the primary stages of the value chain. Although the voluntary, light-touch approach employed by the Guide provides flexibility for stakeholders, this aspect could render it ineffective on the other extreme. With these initial considerations in mind, we examine aspects of the ASEAN AI Guide which hit the mark, as well as areas that require more consideration, and suggest policy recommendations to help make it more well-rounded. Additional measures to heighten its use as a safeguard for stakeholders of AI technologies in its development and usage will also be discussed.