by Tan Aik Keong (AK)
Artificial intelligence's rapid development is reshaping the global economy and society, and a key question comes with it: sovereignty. AI sovereignty refers to a country's control over its own data, infrastructure and applications — a critical condition for protecting data security, fostering innovation and driving economic growth. In Malaysia, that vision closely ties to the work of the Malaysia Digital Economy Corporation (MDEC) and the newly established National AI Office (NAIO). Here's why AI sovereignty matters, what other countries are doing about it, and how Malaysia is building its own AI sovereignty framework through NAIO.
Why AI sovereignty matters
In today's data-driven world, AI sovereignty isn't a luxury — it's a necessity. The concept goes beyond just using AI technology within national borders; it covers control over data storage, processing, and the development of local AI capability. That control matters because AI models typically rely on large volumes of data to make decisions or generate insights. When that data sits outside national borders, it risks unauthorised access or misuse — a real threat to national security and economic stability. A country without control over its own AI systems is vulnerable to outside influence, risking both its competitive edge and its citizens' privacy. For Malaysia, achieving AI sovereignty means building an AI ecosystem suited to its own culture, economy and security needs, strengthening the country's independence in a data-driven future.
What other countries are doing
Around the world, many countries have recognised the need for AI sovereignty and are taking steps to build AI systems suited to their own regional and linguistic character. A few examples:
UAE: through its Falcon model, the UAE has built an AI system for Arabic-language applications, focused on enterprise-grade solutions for sectors like banking and healthcare — helping it meet its own needs without depending on foreign AI technology, keeping control over sensitive data.
China: has developed one of the world's largest language models, Wudao 2.0, making it a serious force in AI development — reducing dependence on foreign AI technology, and the data-security and digital-espionage risk that comes with it.
Russia: Yandex's YaLM model offers a homegrown alternative to international models like GPT-3, purpose-built for Russia's language and regulatory needs — a localised approach that keeps its digital economy on solid footing without relying on outside AI technology.
Singapore: through its SEA-LION model, focuses on supporting Southeast Asian languages, ensuring inclusivity and data privacy while meeting local market needs.
These examples all show how AI sovereignty lets a country protect data security, adapt to its own cultural and linguistic context, and strengthen economic independence. For Malaysia, following a similar path could mean building models that support local languages — Bahasa Malaysia and various indigenous languages — alongside a framework that protects national interests.
Malaysia's own path forward: the National AI Office
Malaysia's vision for AI sovereignty shows up in the strategic establishment of the National AI Office (NAIO), as recent government announcements have detailed. NAIO is meant to shape Malaysia's AI future, encouraging both public and private sectors to adopt AI technology aligned with national priorities. One of NAIO's key functions will be overseeing and managing Malaysia's AI infrastructure and datasets, driving a responsible, secure AI ecosystem. In line with that, related initiatives also emphasise the need for a National Sovereign Large Language Model (NSLLM), focused on Malaysia's specific needs across healthcare, financial services and public services. By developing a homegrown LLM, Malaysia can both secure its data and foster innovation suited to its own socioeconomic context.
Challenges and opportunities ahead
For all the potential benefit, the path forward isn't without real challenges. Key issues include building robust data-governance policy, investing in infrastructure, and cultivating skilled talent. Countries like the UAE and Singapore have addressed these challenges effectively by building collaboration between government, academia and industry — a model Malaysia can draw on too, through NAIO and MDEC, to cultivate talent and keep its AI sovereignty efforts sustainable.
Beyond that, tackling these challenges gives Malaysia the chance to become a regional AI leader — building export-ready solutions suited to Southeast Asia's needs, and contributing to the region's economic resilience.
Closing
AI sovereignty is essential to protecting national data security, fostering local innovation and strengthening economic independence. For Malaysia, initiatives like NAIO and the NSLLM are important milestones on that journey, positioning the country as a secure AI economy for the future. By drawing on other countries' experience and building a framework that supports local development, Malaysia can ensure its AI sovereignty efforts protect national interests and foster a robust digital ecosystem. With careful planning, investment and collaboration, Malaysia can earn its place on the global AI stage, and secure its future in the digital era.
Part of the AK AI Corner column. Originally published in Oriental Daily (东方日报) on Nov 3, 2024.