The phrase “in the cloud” makes computing sound weightless. In reality, cloud services live in buildings filled with servers, cooling systems, backup equipment and network connections, all of which require electricity around the clock.
AI workloads have intensified attention on data-center power because dense computing can create large, relatively steady loads. Utilities may need new substations, transmission upgrades or generation capacity long before a project opens its doors.
Why this matters now
Local officials face a balancing act. Large facilities can bring construction spending and tax revenue, but they may also consume land and infrastructure while creating fewer permanent jobs than other kinds of industrial development.
Water is another variable, though the impact depends heavily on cooling design and climate. Some facilities use evaporative cooling; others rely more on air, closed loops or newer liquid-cooling systems. Broad claims about water use can be misleading without site-level details.
What to watch next
The most useful public debates focus on transparency: expected peak demand, who pays for grid upgrades, backup-generation rules, water plans and whether clean-energy commitments add new supply or merely re-label power that already exists.
Data centers are becoming a planning issue because digital growth is colliding with physical limits. The question is no longer whether computing uses infrastructure; it is how communities want that infrastructure built and paid for.
