Could AI data centers make California’s wildfire problem even worse? Here’s what locals told me
The rapid buildout of AI infrastructure has pushed data centers into the national spotlight as utilities race to meet rising electricity demand. In California, that debate is getting more local as residents in fire-prone regions ask whether large server campuses could add pressure to already strained power and water systems. Interviews with locals this month matched concerns already showing up in utility filings, state planning documents, and company statements.
Data center growth is accelerating, and utilities are planning for it
PG&E told investors in 2024 and 2025 that data centers are a major source of new electricity demand in Northern California, with company executives saying large-load requests have increased sharply. The California Energy Commission has also flagged data centers as part of the state’s broader load-growth picture in recent planning updates. Those documents do not say AI data centers directly cause wildfires, but they do confirm that power demand from computing facilities is rising.
One project residents mentioned repeatedly is the growing concentration of industrial and logistics development in inland California, where heat, drought, and fire weather already shape local planning. Amazon Web Services, Microsoft, Google, and Oracle have all expanded California data operations over time, though companies have not released a single statewide count of AI-specific sites. What is verified is the scale of power these campuses can require, often measured in tens or hundreds of megawatts, according to utility and industry disclosures.
In California, the concern is less the buildings and more the systems around them
Residents in parts of Solano County, the Central Valley, and inland Southern California said their worries center on transmission lines, backup generation, and water demand during extreme heat. Those concerns line up with California’s wildfire history: the state recorded thousands of wildfires in recent years, and CAL FIRE has repeatedly identified power equipment as one of several ignition sources in major fires. A data center itself is not the same as a wildfire ignition point, but the supporting grid infrastructure can matter.
What is not yet known is exactly which proposed or future California AI sites could require new substations, new transmission upgrades, or major water commitments in the highest-risk fire zones. Companies have not released a comprehensive list of all California locations tied specifically to AI expansion. State and local permitting records are also fragmented across counties, which makes statewide comparisons difficult.
Why this issue is coming up now, and what residents should expect next
The main driver is AI demand. Executives from major tech firms including Microsoft and Google have said in earnings calls and public statements since 2024 that AI services require far more computing capacity than earlier products, which means more servers, more cooling, and more electricity. The International Energy Agency and U.S. utility planners have also warned that data center growth is becoming a significant factor in future power forecasts.
For Californians, that means more scrutiny of where large facilities are built and how utilities serve them. Residents should expect environmental review fights to focus on water, backup diesel use, substation upgrades, and transmission routing in counties with elevated fire risk. So far, the verified record shows a risk debate, not proof that AI data centers are making California wildfires worse today, while regulators and utilities continue reviewing new loads and grid upgrades.
