RT AI TOOLKIT

Expanded ASEAN Guide on AI Governance and Ethics

Introduction

 

The rapid development of artificial intelligence (“AI“) shows no sign of slowing down, and with its increased proliferation, governments have been seeking to establish national frameworks to ensure the controlled and secure development and deployment of AI. In recognition of the cross-boundary nature of AI, governments have also been looking towards greater cooperation in AI framework building and standards equivalence.

 

In February 2024, the ASEAN member states issued the ASEAN Guide on AI Governance and Ethics (“ASEAN AI Guide 2024“), which establishes common principles and recommends best practices on the implementation of trustworthy AI in ASEAN. Standing as a landmark regional guide on AI policy, it seeks to help promote consumer confidence and facilitate cross-border deployment of AI services and solutions.

 

The ASEAN AI Guide 2024 has now been updated with the release of the Expanded ASEAN Guide on AI Governance and Ethics – Gen AI (“Gen AI Guide 2025“). The Gen AI Guide supplements and supports the ASEAN AI Guide 2024 with policy considerations related to Gen AI, highlighting the opportunities and risks of Gen AI, and recommending a range of policy actions for ASEAN to support its responsible adoption.

 

The issuance of the Gen AI Guide 2025 demonstrates the continually evolving nature of AI policy, which seeks to address developing technologies and manage emerging risks. It also shows the proactive approach that ASEAN member states are taking to facilitate a regional-level advancement in AI management.

 

This Update provides a summary of the key elements of the Gen AI Guide 2025, including the highlighted Gen AI risks and the policy recommendations.

 

Guiding Principles

 

The Gen AI Guide 2025 is designed as a resource for ASEAN policymakers to understand the primary challenges that they might encounter when working on subjects related to Gen AI, as well as potential strategies for addressing them.

 

The ASEAN AI Guide 2024 set out seven guiding principles for ensuring trust in AI and the design, development, and deployment of ethical AI systems. These principles continue to be relevant in the Gen AI Guide 2025 which highlights additional considerations in the context of Gen AI.

 

The guiding principles in the guides are as follows:

 

  • Transparency and Explainability
  • Fairness and Equity
  • Security and Safety
  • Robustness and Reliability
  • Human-centricity
  • Privacy and Data Governance
  • Accountability and Integrity

 

Gen AI Risks

 

While the ASEAN AI Guide 2024 considers the risks posed by the adoption and development of traditional AI, the Gen AI Guide 2025 recognises that Gen AI may carry risks that are not typically associated with traditional AI, or risks that are enhanced and whose impacts or mitigations may change in the context of Gen AI. These risks, which include ethical, legal, and societal issues,  require new approaches to governance.

 

The Gen AI Guide 2025 thus identifies the following unique Gen AI risks, which form the basis of its policy recommendations:

 

  • Mistakes and anthropomorphism: Gen AI systems can make highly coherent and persuasive mistakes, referred to as “hallucinations”.
  • Factually inaccurate responses and disinformation: Gen AI systems can amplify false or misleading information.
  • Deepfakes, impersonation, fraudulent and malicious activities: Gen AI systems pose risks of impersonation or misinformation by creating realistic content like deepfakes and phishing emails.
  • Infringement of intellectual property rights: Gen AI systems may lead to legal repercussions if copyrighted works are used as data to train the systems without an appropriate legal basis, or if the generated content too closely resembles existing works.
  • Privacy and confidentiality: Gen AI systems may memorise and reproduce specific training data, or otherwise allow malicious actors to reconstruct sensitive information.
  • Propagation of embedded biases: Gen AI systems can inherit and reflect biases from their training data, leading to biased or toxic outputs that reinforce stereotypes.

 

Policy Recommendations

 

The ASEAN AI Guide 2024 provides guidance on measures promoting the responsible use of AI that organisations should adopt. It also provides national-level recommendations to ensure responsible design, development, and deployment of AI systems, and regional-level recommendations for policy makers.

 

In contrast, the Gen AI Guide 2025 only sets out recommendations to regional policymakers at the ASEAN level. Recommendations to national policymakers are not included in this exercise.

 

The recommendations in the Gen AI Guide 2025 include the following:

 

  • Accountability: ASEAN can encourage and facilitate collaboration among developers, deployers, regulators, cloud providers, and civil society to create a common understanding on what shared responsibility would entail.
  • Data: ASEAN can work together to promote the collation of high-quality open datasets and support for industry data sharing. ASEAN can also develop a common approach towards the use and sharing of personal data across ASEAN, as well as data handling, storage, and governance for Gen AI.
  • Trusted Development and Deployment: ASEAN can explore establishing guidelines for the development and deployment of Gen AI models and/or applications, and should be aware of the critical need for upstream developers to share important information with their downstream deployers.
  • Incident Reporting: ASEAN can explore creating a common understanding around incident reporting and incident management.
  • Testing and Assurance: ASEAN can explore developing a regional benchmark that can facilitate testing to address concerns relevant to ASEAN (e.g. qualitative and quantitative indicators relevant to ASEAN’s unique linguistic, cultural, and societal contexts), as well as a list of evaluation tools and techniques preferred for use across the region.
  • Security: ASEAN can explore supporting and coordinating measures to promote vulnerability detection, promoting security knowledge-sharing among AI ecosystem stakeholders, and establishing guidelines on security by design.
  • Content Provenance: ASEAN can explore the adoption of content provenance technologies accompanied by policies and enforcement measures, as well as regional repository of real world examples of techniques and approaches towards establishing content provenance.
  • Safety and Alignment Research & Development: ASEAN can use its role to promote or convene platforms to facilitate the sharing of insights on AI safety and alignment research across member states and to help identify regional priorities for areas of further research.
  • AI for Public Good: ASEAN can create a regional compendium of responsible Gen AI use cases. It can also promote awareness and education on Gen AI, including skill development, digital literacy and risk awareness, and educational initiatives for workforces.

 

Concluding Words

 

The Gen AI Guide 2025 seeks to highlight that the new challenges of Gen AI do not outweigh the substantial opportunities to improve lives, create economic opportunity, and empower new groups in society. It provides a policy-focused view of the risks and recommendations to promote responsible, risk-aware Gen AI use that ASEAN as a region should consider addressing. Importantly, the Gen AI Guide 2025 sets out the distinction between the risks and considerations between Gen AI and traditional AI, and delves deeper into the unique measures and approaches that are appropriate to Gen AI.

 

For further details, the ASEAN AI Guide 2024 is available here, and the Gen AI Guide 2025 is available here. Please feel free to contact our team for further queries.