Arcondec editorial · 12 / 09 / 2026

AI and power: electricity demand rises

The IEA examines computing’s pressure on energy infrastructure.

The International Energy Agency reported that data center electricity consumption rose about 17% in 2025, compared with 3% for worldwide demand. Its April 2026 update examines the relationship between artificial intelligence and energy. The report’s projections describe future scenarios, not capacity already installed.

Power also shapes the schedule

Data center development needs to consider grid connections and equipment availability alongside building construction. Early coordination across disciplines can reveal dependencies. For Mexico, the news provides global context for discussions of digital infrastructure without replacing project-specific studies.

Editorial analysis

Distinguishing capacity, consumption and growth

When reading industry announcements, it matters what is being measured. Planned capacity, observed consumption and demand forecasts describe different things. Combining them can lead to incorrect conclusions about a development’s scale or progress. The figures cited in this story describe a global consumption change; they cannot directly establish demand for a Mexican city or a new campus.

Coordinating digital and physical timelines

Our editorial view is that technological expansion should be discussed alongside the infrastructure timelines that make it possible. A programme can organise decisions about buildings, equipment, services and operations while making their dependencies explicit. The question is not only how much capacity is desired, but what must be available before each phase. Presenting a project this way makes assumptions easier to explain and schedules easier to reassess when those assumptions change.

What this means for discussions in Mexico

International context helps frame local questions, but it does not answer them automatically. Each initiative needs to establish its location, scale, expected demand and operating requirements. To follow the industry carefully, we suggest comparing figures with equivalent scope and distinguishing estimates from measurements. This allows discussion of artificial intelligence to move beyond headline numbers towards planning decisions that make sense for a particular project.