Trend AI take-off and unprecedented data-centre build-out
Summary
The report sees AI development taking off sharply, with the big technology firms competing to become the largest and most successful players in generative AI. Agentic AI, which plans on its own and makes decisions and carries out tasks with minimal human involvement, is highlighted as a key development. This race channels huge sums into AI infrastructure, and data centres are going up on an unprecedented scale to supply the vast computing power AI needs. Data centres consume very large amounts of energy and demand major investment, and capital spending in the sector has almost doubled since early 2025. The report regards the location of computing capacity, energy and data centres as a matter of power and of where activity settles. In its introduction it judges that the exponential rise in compute and usage and the billions flowing into firms, models and data centres carry consequences for both competitiveness and security.
Classification
Evidence 1
- Rijksoverheid — Ministerie van Binnenlandse Zaken en Koninkrijksrelaties (BZK), Directie KIEM, namens de werkgroep Rijksbrede Trendverkenning Rijksbrede trendverkenning Rijksoverheid — Ministerie van Binnenlandse Zaken en Koninkrijksrelaties (BZK), Directie KIEM, namens de werkgroep Rijksbrede Trendverkenning no link — bibliographic entry pp. 6, 32, 34 2026-05 accessed 2026-09-30
Observed signals 6
- SignalAI sector capital expenditure nearly doubled since early 2025
- SignalData-centre electricity demand could double by 2030
- SignalGujarat launches 'Viksit Gujarat – Data Centre Policy 2026-29'
- SignalReflection AI begins $6.3B compute lease at SpaceX Colossus 2 (Memphis)
- SignalThailand establishes National Semiconductor Policy Committee, approves $20B in digital projects
- SignalUS Eases Funding Rules for Data Centers, Exempts AI-Backed Securities
Part of issues 2
- IssueRisk of an AI investment bubble bursting1 trends · 1 signals
- IssueThe hidden costs of AI: energy, water, minerals and invisible labour1 trends · 0 signals
Relation types: constitutes · supports
Public id: fm-3cfe3eb16248
