The evolution of Digital Public Infrastructure (DPI) is shifting from rigid technical architectures toward inclusive, citizen-centric ecosystems. At the Festival of Innovation 2026, global leaders emphasized that digitizing services is not synonymous with true digital transformation if it fails to address the needs of the excluded. To bridge this gap, experts suggest a co-creation model. While foundational layers like system architecture remain centralized, the application layer—where citizens directly interact with services—should be designed from the bottom up to ensure real-world usability and trust.

Interoperability (the ability for different systems to work together) remains a significant hurdle, often viewed solely through a technical lens. However, a robust DPI requires a four-layered approach. This begins with legal mandates to share data, followed by aligned workflows across agencies (organisational interoperability), and a shared understanding of data meaning (semantic interoperability). By establishing security by design and privacy by design as foundational principles, governments can mitigate the risks of data exposure while building systems that evolve safely alongside societal needs.

The build-versus-buy debate for national AI is finding a middle ground through open-source adoption. For regions like Southeast Asia, where local cultural and linguistic contexts are underrepresented in global datasets, open-source AI serves as a strategic tool for digital sovereignty. By adapting global models to local needs, governments can maintain technological relevance without sacrificing their unique regional perspectives. Education and hybrid low-tech options, such as QR-coded paper vouchers for the elderly, remain essential to prevent new forms of digital exclusion.

The evolution of Digital Public Infrastructure (DPI) is shifting from rigid technical architectures toward inclusive, citizen-centric ecosystems. At the Festival of Innovation 2026, global leaders emphasized that digitizing services is not synonymous with true digital transformation if it fails to address the needs of the excluded. To bridge this gap, experts suggest a co-creation model. While foundational layers like system architecture remain centralized, the application layer—where citizens directly interact with services—should be designed from the bottom up to ensure real-world usability and trust.

Interoperability (the ability for different systems to work together) remains a significant hurdle, often viewed solely through a technical lens. However, a robust DPI requires a four-layered approach. This begins with legal mandates to share data, followed by aligned workflows across agencies (organisational interoperability), and a shared understanding of data meaning (semantic interoperability). By establishing security by design and privacy by design as foundational principles, governments can mitigate the risks of data exposure while building systems that evolve safely alongside societal needs.

The build-versus-buy debate for national AI is finding a middle ground through open-source adoption. For regions like Southeast Asia, where local cultural and linguistic contexts are underrepresented in global datasets, open-source AI serves as a strategic tool for digital sovereignty. By adapting global models to local needs, governments can maintain technological relevance without sacrificing their unique regional perspectives. Education and hybrid low-tech options, such as QR-coded paper vouchers for the elderly, remain essential to prevent new forms of digital exclusion.