WEF-based analysis highlights power, land and connectivity constraints on AI data-centre growth
DataNews Editorial Team
Data Centre Magazine has published statistics drawn from the World Economic Forum’s AI infrastructure decision playbook, outlining the projected scale of global data-centre expansion through 2030. The figures point to potential US$7 trillion investment alongside rapidly rising requirements for electricity, developable land and interconnect capacity.
- The cited WEF-based outlook puts potential global data-centre investment at US$7 trillion by 2030.
- Data-centre electricity demand is projected to rise by more than 20% annually through 2030.
- An additional 160 km² of land may be required for data-centre site expansion by 2030.
- AI could increase data-centre interconnect bandwidth demand sixfold over the next five years.
- Ant Group reported that clean electricity represented approximately 65% of its data-centre electricity consumption in 2025.
Data Centre Magazine’s third-week-of-October statistics feature reviewed the World Economic Forum report, “Data Centre Sustainability and Resilience: A Decision Playbook for AI Infrastructure.” The article consolidates projections for investment, land demand, electricity consumption and interconnection bandwidth as AI workloads drive the next phase of data-centre development.
The cited outlook places potential global data-centre investment at US$7 trillion by 2030. It also indicates that data-centre electricity demand is projected to grow by more than 20% annually through 2030. For operators developing AI capacity, this makes power availability and grid interconnection central considerations rather than downstream delivery issues.
Land requirements are also expected to tighten. The article cites 81 km² as the area of powered land supporting data centres worldwide and estimates that a further 160 km² of land may be needed to support site expansion by 2030. Availability of both suitable land and power infrastructure is characterised as a principal constraint on further sector growth.
Connectivity is a parallel limitation. AI could drive a sixfold increase in demand for data-centre interconnect bandwidth over the next five years, according to the figures cited. This could increase the importance of network capacity alongside power and site readiness when evaluating markets for AI-oriented campuses.
The article also references Ant Group’s reported energy mix: clean electricity accounted for approximately 65% of its data-centre electricity use in 2025. This is a company-specific disclosure, rather than an industry benchmark, but illustrates the growing relevance of electricity sourcing and decarbonisation strategies in data-centre planning.
Taken together, the projections suggest that competition for available generation, grid access, connected land and interconnection capacity could intensify faster than traditional data-centre development cycles. The underlying estimates remain forward-looking, however, and should not be interpreted as committed investment, completed capacity or guaranteed infrastructure delivery.
The analysis reinforces that AI-oriented data-centre expansion is increasingly constrained by the combined availability of power, grid interconnection, suitable sites and network connectivity. If the cited projections materialise, operators and investors may need to treat these factors as interdependent constraints in market selection and capacity planning.
- Data Centre Magazine / October 25, 2026 · EnglishSource →
This article was prepared by the DataNews editorial team based on the sources listed above.