As artificial intelligence adoption accelerates across Asia, one challenge continues to stand out: trust. A new report by the Asia Society Policy Institute (ASPI), Designing Metrics to Enable Trusted AI Ecosystems in Asia, argues that Asia needs common frameworks and measurable indicators to ensure AI systems are safe, reliable, and inclusive.
The paper identifies nine key pillars for trusted AI ecosystems, including trusted datasets, AI infrastructure, AI skills and awareness, ethical AI development, governance frameworks, misinformation management, cybersecurity, and environmental sustainability. Rather than focusing only on regulation, the report emphasizes the importance of measurable trust indicators that can help governments and institutions benchmark progress and identify gaps.
A major takeaway from the report is that Asia’s AI landscape is highly diverse, with countries at different stages of digital maturity and governance readiness. The authors argue that while a single governance model may not fit every country, a shared metrics framework can help create interoperability, transparency, and regional cooperation.
AI Asia Pacific Institute was pleased to contribute to these discussions through participation in the Asia Society Policy Institute roundtables held during the India AI Impact Summit 2026 in New Delhi. These conversations brought together policymakers, industry leaders, civil society, and international organizations to explore practical pathways for enabling trustworthy AI ecosystems across the region. AIAPI suggested an AI Technical Facility in these discussions.
The report’s recommendations strongly align with the mission of the Asia Knowledge Facility, which supports regional collaboration, knowledge-sharing, and evidence-based policymaking on emerging technologies and AI governance in Asia. As AI adoption grows, initiatives that strengthen trust, accountability, and cross-border cooperation will be essential to ensuring AI benefits communities throughout the region.
Read more on our ASEAN discussion paper here.
