AI has gained global attention, but concerns have been raised about its control by non-state actors and its impact on citizens. While most AI experts agree that global efforts are needed to promote AI, China’s Global AI Governance Initiative and the US tightening export controls over advanced computing chips raise questions about the effectiveness of multilateral efforts. Regional coordination of AI governance is crucial, especially in Asia, where reforming the service sector to harness the digital revolution is essential for inclusive and sustainable growth. Coordinated regional arrangements can mitigate geostrategic competition between the US and China while reducing middle powers’ need to choose sides.
However, effective AI governance faces challenges such as the concentration of power over AI inputs by the US, China, and a few technology infrastructure firms, and the exclusion of women, rural residents, and indigenous populations from accessing AI benefits. Despite differences in state perspectives and capabilities, the region has the necessary ingredients to shape a regional framework for AI governance, including flexible digital policy tools and industry engagement strategies.
AI governance in Asia faces a challenge due to the dominance of US and Chinese technology infrastructure companies over key inputs, particularly large language models (LLMs), which rely on data and computation-intensive machine learning. This concentration of power makes it difficult for new entrants to compete and public actors to ensure transparency and accountability of AI systems. Some governments across the Asia Pacific are seeking to protect and localize their digital assets through national policy, which has negative impacts on AI systems, such as reduced access to training data, starved innovation ecosystems, and fragmentation of cybersecurity mechanisms.
The US is actively promoting data localisation through the RCEP trade agreement, investing in onshore GPU production, AI innovation, and export controls targeting high-end GPUs sold to China. Without a robust regional framework to counteract localisation, potential AI competitors like China, India, and Indonesia may respond in kind, leaving smaller and poorer countries with fewer options to participate in the AI industry. Southeast Asia’s weak AI readiness risks the region’s digital divides becoming ‘algorithmic divides’, with an estimated 61% of ASEAN populations not using the internet despite living within range of internet access.
To address concentration, new paradigms of data ownership and valuation are needed, including data cooperatives and data unions. Capital providers can support the development of SME- and community-driven AI systems, reducing reliance on large-scale proprietary models and centralized cloud computing infrastructure. Regional coordination of third-party AI oversight can lower the costs of regulation at the national level. Singapore’s AI Verify Foundation and a proposed global regulatory sandbox initiative could begin in Asia.
Counterbalancing localisation can be achieved by updating bilateral, minilateral, and multilateral trade agreements for cross-border data flows. Examining national security exemptions in multilateral trade rules can help distinguish which AI-relevant assets could be liberalized. A regional standards body can ensure accountability without compromise, while regulatory leaders can strengthen regulations and AI strategies with ASEAN and Pacific Island nations. Equitable participation in regional AI ecosystems requires SME financing and digital capacity building.