In the California farming town of Gilroy — best known as the self-proclaimed garlic capital of the world — construction cranes appeared one day on a 56-acre lot, signaling the arrival of a massive $2 billion data center. The problem? Almost no one in the community knew it was coming, and nobody had voted on it. Nobody had to.
Amazon Web Services quietly secured approval for the 438,500-square-foot AI data center through staff-level permits that leveraged decades-old industrial zoning, entirely bypassing the public hearings and City Council votes that residents might have expected. For ongoing coverage of the AI industry's infrastructure buildout, AI Buzz Wire is following this story closely.
How Amazon Bypassed Public Scrutiny
The mechanism that enabled the project is striking in its simplicity. AWS purchased the parcel at 8050 Camino Arroyo in 2020 and negotiated with city officials for several years. On July 3, 2023, Gilroy's community development director signed off on the facility using an administrative approval process — no City Council vote, no Planning Commission hearing, and according to official city project materials, no public hearing required.
The key was zoning. Gilroy's land-use framework dates back decades, and the lot had been tagged as industrial long before anyone imagined a hyperscale cloud campus descending on an agricultural community. Under those rules, a data center qualified as an industrial use and could be approved at the staff level — the same channel used for warehouses and food processing plants.
The rules were followed perfectly. That is precisely what makes the situation uncomfortable for residents who feel they were shut out of a decision that will reshape their community for decades.
Scale Changes Everything
Mayor Greg Bozzo acknowledged that prior industrial projects moved through the same approval channel without controversy. But a $2 billion, 438,500-square-foot data center is not a garlic dehydration facility. The sheer scale of the operation — and its voracious demand for electricity and water — places it in an entirely different category from the industrial uses the zoning framework was designed to accommodate.
Gilroy sits in an agricultural region where water is already a charged issue. Residents have zeroed in on the data center's projected water consumption, alongside concerns about fire risk from battery storage systems and the structural impossibility of providing meaningful community input after construction had already begun.
Their objection is not that Amazon broke the law. It is that the law never required their participation in the first place.
A National Pattern
Gilroy's experience is not an isolated case. It previews what dozens of communities with legacy industrial zones across the United States may soon face as the AI data center boom accelerates.
Hyperscale data centers are landing in municipalities nationwide through identical staff-level permitting channels. The mechanism that Amazon used in Gilroy is available in most towns with old industrial zoning — and most towns have one. As tech giants race to build AI infrastructure, the gap between community awareness and corporate construction timelines is widening.
According to Wall Street Journal reporting, Amazon's Roger Wehner acknowledged the communication gap and said the company recognized the community wanted earlier engagement. AWS has since hosted an open house for residents, though critics note this came well after the project was already under construction.
The Broader AI Infrastructure Surge
The Gilroy data center is part of an unprecedented wave of AI infrastructure investment. Driven by surging demand for cloud computing power to train and run large language models, hyperscale operators are projected to spend hundreds of billions of dollars on data center construction through the end of the decade. Amazon alone has hiked its 2026 capital expenditure plans to over $220 billion, citing an AWS AI surge.
But the Gilroy story exposes a critical blind spot in this buildout: the local governance frameworks designed to give communities a voice in major development decisions were written long before the era of hyperscale AI infrastructure. Zoning codes drafted in the 1970s and 1980s never contemplated facilities that consume as much electricity as a small city and draw millions of gallons of water for cooling.
The result is a structural mismatch. Companies can move at the speed of cloud computing while communities are locked into permitting processes designed for a different era.
Gilroy Fights Back
In the wake of the controversy, Gilroy officials are rewriting their zoning codes to require Planning Commission review for future data center projects — closing the loophole that allowed Amazon's facility to sail through without public input.
The fix, while meaningful for future projects, comes too late to alter the course of the $2 billion facility already rising on Camino Arroyo. For residents who learned about the data center by accident, the zoning overhaul represents a hard-earned lesson in civic vigilance.
The Gilroy case is likely to become a template — both for communities seeking to assert control over AI infrastructure siting and for tech companies evaluating which municipalities offer the path of least resistance. As the data center gold rush continues, the battle over who gets a say in shaping the physical footprint of AI is only beginning.
Stay Ahead of AI
For more reporting on AI infrastructure, data center policy, and the tech industry's expanding footprint, follow AI Buzz Wire for daily coverage.
Read more AI news →
