Southeast Asian nations are pursuing a “third pathway” to establish technological autonomy in the artificial intelligence race between China and the United States, according to an ISEAS Perspective paper published in October by John Lee. Rather than accepting complete integration into either American or Chinese tech stacks, countries like Vietnam, Singapore, Malaysia, Indonesia, and Thailand are adapting Cold War-era political hedging strategies to balance great power competition and protect domestic data sovereignty.
The Great Power Tech Stack Race in Southeast Asia
Technological competition between Washington and Beijing mirrors traditional diplomatic and security rivalries across Southeast Asia. According to policy analyses, the United States promotes its full hardware and software ecosystem through initiatives like the American AI Exports Program, pushing American AI infrastructure as the global gold standard while some policymakers debate an AI Marshall Plan. Meanwhile, China invests heavily in domestic stack capabilities to replace Western dominance and shape regional international standards. Refusing to remain mere consumers, Southeast Asian governments are exploring domestic stack development and localized policy frameworks to secure independence over digital data streams.
Vietnam’s Sovereign Cloud and Localized AI Models
Vietnam has taken concrete legislative steps by introducing Southeast Asia’s first law on AI alongside the Law on the Digital Technology Industry, as outlined in regional policy reviews. Hanoi classifies digital systems as strategic assets subject to strict data localization mandates. State-linked firm FPT anchors its compute capacity in the FPT AI Factory utilizing US Nvidia chips while relying on a sovereign cloud known as FPT Smart Cloud. Private enterprises complement this by building native language models like PhoGPT, developed by VinAI, and Semikong, which was built with FPT software support. These localized large language models are engineered to handle regional contexts and reduce reliance on foreign AI systems.
Did you know? Vietnam’s national AI strategy relies on asymmetrical scaling and adaptation, combining foreign hardware like Nvidia chips with sovereign cloud infrastructure to avoid total technological dependency.
Singapore and Regional Hedging Initiatives
Singapore leverages public policy to maintain domestic oversight of its AI ecosystems, launching the Model AI Governance Framework for Agentic AI in January to update prior generative AI policies. State-backed initiatives such as AI Singapore developed the SEA-LION multimodal open-source model family specifically to counter the under-representation of Southeast Asian languages and cultural contexts in Western and Chinese frontier models. Neighboring countries including Malaysia, Indonesia, and Thailand similarly hedge between US and Chinese investments to retain control over their domestic systems.
Why National Hedging Faces Hurdles
National AI strategies focused solely on individual hedging are only partially successful because local efforts cannot easily match foreign resources. According to analysts Jacob Taylor and Joshua Tan, national initiatives are simply no match for “the scale of compute, talent or high-quality data of US or Chinese AI ecosystems.” Because regional firms remain dependent on American and Chinese software and hardware, efforts to build sovereign stacks often result in localized adaptations of foreign models backed by technology transfers rather than entirely independent domestic technologies.
Collaborative Hedging and the “Airbus for ASEAN” Concept
To overcome individual limitations, researchers suggest that Southeast Asian nations pivot toward collaborative hedging rather than isolated national strategies. Taylor and Tan propose adapting the Airbus consortium model as a template for regional technological cooperation. This approach would establish a public-private network of middle-power national AI laboratories tasked with building scalable AI products under a public utility framework. Collective diplomatic alignment helps ensure that ASEAN members maintain consistent infrastructure standards without aligning ideologically with either superpower.
Frequently Asked Questions
-
What is the “third stack” movement in Southeast Asia?
It is an initiative by Southeast Asian nations to build domestic AI capabilities, deploy local policy frameworks, and reduce total reliance on American or Chinese technology products. -
How is Vietnam working toward AI sovereignty?
According to regional policy documents, Vietnam implements localized data laws, utilizes sovereign cloud platforms like FPT Smart Cloud, and develops native large language models such as PhoGPT. -
What is collaborative hedging in AI development?
Collaborative hedging involves regional coordination among middle powers—similar to public-private industrial consortia—to pool compute resources, talent, and data standards outside the direct control of Washington or Beijing.
What are your thoughts on Southeast Asia’s push for AI autonomy? Share your perspective in the comments below, explore our latest tech policy analyses, and subscribe to our newsletter for ongoing updates on global technology competition.
Related reading