Mount Shasta City Manager Discloses Inquiry for AI Data Center at Former Crystal Geyser Site

Key Takeaways

• The Mount Shasta city manager confirmed an inquiry about an AI data center at the former Crystal Geyser facility.
• Community concerns center on water use, energy demand, and land-use control as Northern California faces drought conditions.
• The situation reflects broader national trends in rural data center siting and associated governance tensions.

The conversation over a possible AI data center near Mount Shasta is drawing intense local interest, partly because the location carries heavy historical context. When the Mount Shasta city manager revealed that the city had received an inquiry about using the former Crystal Geyser bottling site, it landed in a community that has spent years debating how large industrial projects align with local identity. The site itself is in Siskiyou County rather than within city limits, which complicates the permitting picture.

No one yet knows which entity approached the city. The planning director received the inquiry recently, and the city manager has kept the name confidential. Even without that detail, the announcement has sparked fresh debate at a moment when California is again grappling with drought and long-term water reliability.

External research shows why the reaction was intense. According to the Pew Research Center, more than 3,000 data centers operate in the United States and over 1,500 more are in development. Many of the new builds target rural areas, where land tends to be cheaper and local zoning processes can be more flexible. Mount Shasta fits that pattern. With a population around 3,000, the city also occupies a unique environmental position as a key contributor to Northern California’s water resources.

During the city council meeting, the city manager stated openly that he was skeptical of the fit, noting he could not imagine the project benefiting the community unless it offered operations substantially different from typical data center models. Water consumption was at the center of that concern. Modern AI and high-performance computing facilities require tens of megawatts of constant electricity and robust cooling infrastructure. The NIST SP 500 series notes that these environments tend to place heavy stress on local water and grid systems, especially when designed for dense model training workloads.

Mount Shasta’s water comes from glaciers and volcanic aquifers, filtering through lava tubes for years before emerging as some of the cleanest spring water in the country. Headwater Springs, where residents and tourists collect water, feeds the Sacramento River system. That makes the region incredibly sensitive to anything that could alter local hydrology, even indirectly.

Local history also plays a major role. When the Crystal Geyser Water Company purchased the site in 2013, the plan to operate a bottling facility spurred years of opposition, much of it led by the Winnemem Wintu Tribe. The company withdrew in 2021. As soon as discussion of an AI data center surfaced, the Tribe publicly reiterated concerns, citing water use, electrical demand, noise, and potential emissions. On social media, their commentary drew wide engagement and helped energize online petitions.

This situation raises a broader question: how should rural communities navigate economic development pressures tied to AI infrastructure? Smaller counties often face asymmetric information. They must evaluate highly technical proposals, balance community priorities, and predict long-term infrastructure needs, all while competing with other regions for investment.

Governance and permitting present additional complexities. Because the former Crystal Geyser property sits in unincorporated Siskiyou County, the county holds the permitting authority. The city manager acknowledged that annexation into the city could shift that balance, giving Mount Shasta greater control over water and land-use decisions. At least one county supervisor has indicated openness to that idea, although there are no formal steps yet.

Meanwhile, residents in nearby areas are confronting similar questions. In Imperial Valley, residents organized to oppose a proposed 950,000-square-foot AI data center over environmental and governance issues. Across the state, there are 45 operating data centers and 27 planned, representing a total capacity of 2,720 megawatts, according to Cleanview 2025. Real-world vendors driving this expansion include hyperscale and colocation providers such as Iron Mountain Data Centers, Vantage Data Centers, and Terra Ventures, all active in large-scale California projects. Evaluating infrastructure investments from major operators like Iron Mountain Data Centers highlights the mounting pressure to find suitable sites.

Mount Shasta is now navigating these state-wide pressures, even if the current inquiry turns out to be preliminary. The city manager emphasized that his early reaction is that such a project would not align with the city’s interests. He pointed to the alluvial plains and long percolation pathways that create the region’s water quality, suggesting that protecting those resources remains a top priority for city leadership. The council does not appear inclined to promote a data center project at this stage.

The next step is a meeting between the unnamed entity and city staff. Community members are preparing to engage as well, with the Siskiyou County Board of Supervisors and the Mount Shasta Council expected to review the matter. Whether the inquiry becomes a formal proposal is still unclear. Yet the debate reveals how quickly an initial inquiry can ignite broader concerns when AI infrastructure intersects with sensitive environmental regions.

Rural communities across the United States are watching these developments closely, and Mount Shasta’s response may become one of several examples shaping how local governments evaluate the next wave of data center growth.