\n\n
Key developments: A new assessment shows Iberia’s data‑center pipeline approaching 9 GW, underscoring the scale of AI‑computing demand in Spain and Portugal. At the same time, community resistance to power‑intensive facilities is accelerating worldwide. In Europe, more than 70 projects were rejected or restricted in just the first four months of 2026, outpacing the entire previous year, while investors worry about securing reliable electricity for the planned capacity. The trend is echoed in Asia, where CoreWeave announced a 240 MW AI campus in Navi Mumbai, expanding its footprint into a market wary of grid strain, and in the United States, Vultr’s $1.2 billion order for HPE‑AMD AI racks signals continued appetite for high‑density compute.
Why it matters: The Iberian pipeline illustrates how AI workloads are driving megawatt‑scale data‑center builds, but without firm power commitments the announced capacity could remain speculative. Growing public opposition—highlighted by the wave of European project rejections—adds regulatory risk and could slow deployment, raising construction costs and delaying revenue streams. Meanwhile, new investments in emerging markets (India) and sizable hardware orders (US) show that the demand side remains robust, creating a tension between supply needs and local acceptance.
What to watch next: • Utility commitments and construction milestones for Iberian sites – will developers lock in grid capacity before the pipeline moves from plan to operation?
• Policy shifts in Europe and Asia – new zoning rules or renewable‑energy mandates could reshape project economics.
• Further large‑scale AI campus announcements – especially in regions with tight power grids, which may prompt innovative cooling or on‑site generation solutions.
• Market response to the $1.2 billion AI rack order – will other hyperscalers follow suit, amplifying the power‑demand narrative?
Stakeholders should monitor utility negotiations, regulatory updates, and the rollout pace of high‑density AI infrastructure to gauge whether the AI‑driven data‑center boom will accelerate or encounter a slowdown due to community and power‑supply constraints. [[1, 2]]