Homegrown AI in Malaysia: Leading ASEAN with Trustworthy Values
- What: Malaysia is developing two AI initiatives - ILMU and NurAI - focused on sovereign capability and ethical, syariah-aligned guidance.
- When: ILMU launched recently by YTL AI Labs Sdn Bhd; NurAI developed by Zetrix AI Bhd.
- Why it Matters: Positions Malaysia as a potential leader in trustworthy, values-based AI, addressing data sovereignty and ethical concerns.
“`html
Malaysia Positions Itself as ASEAN Leader in Trustworthy AI with ILMU and NurAI
Table of Contents
Updated August 25, 2025, 00:58:00
The Rise of Malaysian AI Initiatives
Malaysia is actively pursuing the advancement of its own artificial intelligence (AI) capabilities with two key initiatives: ILMU (Luhur Malaysia Interest For You or Excellent Malaysian Intelligence for You) and NurAI. These projects aim to establish Malaysia as a leading force in trustworthy and values-based AI within the Association of Southeast Asian Nations (ASEAN). According to AI risks and safety adviser Chan Tzu Kit, these initiatives represent a notable step forward for the nation.
ILMU, launched by YTL AI Labs Sdn Bhd, is Malaysia’s first homegrown large language model (LLM). It’s designed as a multimodal model specifically tuned for the Malaysian context, targeting applications in government, education, and small and medium-sized enterprises (SMEs). NurAI, developed by Zetrix AI Bhd, distinguishes itself as the world’s first syariah-aligned, Islamic values-based LLM. This addresses a critical need for curbing misinformation and expanding services within Islamic finance, halal industries, and public guidance.
“Sovereign, locally trained multimodal models also mean sensitive Malaysian data doesn’t have to leave the country,” Chan Tzu Kit stated in an interview with Sunbie. This data sovereignty aspect is a key driver behind the development of ILMU.
from AI User to AI Maker: A Paradigm Shift
Chan described Malaysia’s move towards developing its own AI models as a “step-change,” transitioning the country from being primarily an AI user to an AI maker. This shift signifies a strategic investment in technological independence and innovation. Though, he emphasized that the success of these initiatives will depend heavily on establishing trust through robust safety and governance frameworks.
The development of these llms isn’t simply about technological advancement; it’s about shaping AI to align with Malaysian values and priorities. NurAI, in particular, highlights this focus, aiming to provide AI solutions that are consistent with Islamic principles. This is a unique approach in the global AI landscape.
The Importance of Transparency and Safety
Chan Tzu kit stressed the critical need for transparency and rigorous testing to ensure the reliability and safety of both ILMU and NurAI. He specifically called for the publication of “model cards” – extensive documentation detailing the model’s capabilities,limitations,and potential biases – for both initiatives. These model cards are crucial for fostering accountability and enabling informed decision-making.
he also advocated for independent evaluations, particularly focusing on the models’ performance in Bahasa Melayu, their refusal rates (the frequency with which they decline to answer potentially harmful or inappropriate queries), and their factuality in sensitive areas like religion, health, and finance. Independent evaluation is vital to identify and mitigate potential risks.
Further recommendations included:
- Utilizing authenticated sources and incorporating human oversight for sensitive queries.
- Conducting local “red-teaming” exercises with bug bounties to proactively identify and address vulnerabilities, including potential jailbreaks and data leaks.
- Implementing