VState of Play: Preparedness for the Connected Future JULY 2025 Introduction Remaining competitive and agile in any industry or sector requires scalable and future-ready digital capabilities to facilitate the development and delivery of products and services. Digital public infrastructure (DPI) – including elements such as digital identity, data processing and exchange, and digital ownership – provides a foundation upon which these capabilities can be developed, scaled and innovated. While maintaining DPI systems is often a public sector project, innovation and new system development are largely driven by the private sector, fueled by commercial and consumer demand for emerging technologies such as artificial intelligence (AI), extended reality (XR) and biometrics in an advancing digital landscape. Trustworthiness, interoperability and utility of next-generation products and services built on DPI must be continually evaluated, but prioritizing DPI development is essential to enable the economic and societal benefits of emerging technologies. DPI must be effectively developed and responsibly adopted so that the future of digital interchange, including the agentic web, can be delivered at scale and remains open, accessible and democratic, with sustainable infrastructure for communication, commercial activity and societal interaction. Data exchange and processing At the most foundational level of the digital tech stack is the physical infrastructure that powers, scales and drives processing for all downstream services. Since its inception, the internet’s physical infrastructure has grown into a vast, layered system designed to support global connectivity and data exchange. At its core are transoceanic subsea cables, which currently carry 95% of international data.1 On land, terrestrial fiber networks connect cities and regions, feeding into data centres that process, store and route vast volumes of information. These data centres – ranging from small server rooms to massive hyperscale facilities – have historically been built for general-purpose computing, with standard air cooling and modest power densities. Wireless towers, local exchanges and undersea landing stations have also extended the internet’s reach, enabling broadband, mobile data and cloud services to scale globally. Traditionally, infrastructure development was shaped by the needs of web hosting, video streaming, cloud storage and enterprise computing, all of which drove incremental advances in speed, capacity and geographic coverage. Today, however, the physical infrastructure supporting digital services is rapidly evolving to meet the demands of AI, which requires significantly more compute power, electrical capacity and low-latency data delivery than traditional web services. McKinsey estimates that demand for data centre capacity will increase by ~20% per year from 2023-2030.2 To meet this demand, data centre design and construction must address how to scale with higher efficiency and lower environmental impact. Alongside these, global subsea cable networks remain the backbone of international data transfer, now being upgraded for higher bandwidth and improved resilience to accommodate AI-driven traffic surges. Low Earth orbit (LEO) satellites are emerging as critical complements, extending internet access to remote regions and enabling faster, more reliable connections for real-time AI applications. Together with terrestrial fiber networks and edge computing nodes, these components form a more distributed, energy-aware and latency-optimized architecture that reflects the growing centrality of AI in global digital infrastructure. Forecast Recent global events have highlighted the vulnerabilities of the internet’s physical infrastructure in two key areas. First, data centres are facing increased energy and performance demands by the scaling of AI and XR, exacerbated by mechanical overcrowding, inefficient cooling and outdated hardware. This demand can be answered with more design-efficient approaches and technology adoption, aided by public oversight, development support, monitoring and guidance on good practices. Data centres should consider the need for real-time data synchronization, spatial data privacy and contextual integrity safeguards, and interoperability standards for cross-platform data processing. Without these, as well as more distributed forms of hardware that support edge computing, scaling emerging technologies such as XR and AI will be severely limited. C4AIE Digital Technologies

State of Play 2025