The World Economic Forum (WEF) notes that the further expansion of artificial intelligence (AI) will be constrained not by computing capacity, but by the availability of energy, water, critical materials and land. According to the Forum’s estimates, capital expenditure by major technology companies on AI infrastructure will increase from USD 410 billion in 2025 to USD 700 billion in 2026, while global energy consumption by data centers could reach 945 TWh by 2030. This exceeds the current combined electricity consumption of countries such as Germany and France. At the same time, water consumption by data centers is also expected to increase significantly by 2030, while 43% of the world’s data centers are already located in water-stressed regions. WEF experts also warn of risks related to shortages of copper, lithium and rare earth metals. For example, demand for lithium could increase fivefold by 2030, while demand for copper could rise by around 30–40% in the long term. In this context, the WEF calls for a shift from a narrow technology-focused approach to comprehensive planning for AI infrastructure development, taking into account constraints related to energy, water resources, raw material extraction and environmental sustainability.

Digital transformation of public administration is often focused primarily on speed and efficiency, while issues of transparency, inclusiveness and citizen trust are not sufficiently taken into account. WEF experts note that many digital projects fail to achieve the expected results due to the dominance of a technology-centric approach, in which systems are designed around technologies and institutional constraints rather than real-life situations and citizens’ needs. As a solution, the WEF proposes a set of principles for the responsible implementation of GovTech and digital public service infrastructure. These include a citizen-first approach, digital inclusion, transparency of systems using artificial intelligence, and ensuring digital sovereignty over data. The authors also recommend using modular architectures, open standards, independent audits of AI solutions and permanent mechanisms for citizen feedback. Particular attention is paid to the development of experimental regulatory regimes that allow innovations to be tested without creating risks for critical public services.