How Nashville Used Eminent Domain to Block a Data Center Near Its Zoo

Nashville uses eminent domain to block a data center near its zoo, highlighting AI expansion vs. community conflict.
Nashville's unprecedented use of eminent domain to prevent data center construction near its zoo signals a turning point in how local governments respond to AI infrastructure expansion. As tech giants pour hundreds of billions into data centers, communities nationwide push back over power grid strain, water consumption, noise, and limited job creation, forcing the industry to rethink its social contract with host communities.
A Battle Over Land
As the AI boom drives data center construction to unprecedented heights, Nashville's municipal government has invoked eminent domain to block the construction of a data center near its zoo — an event that is drawing widespread attention. It reflects an increasingly prominent social issue: when tech giants' compute expansion collides with a local community's public interest, which way should the scales tip?
Eminent domain refers to the government's legal power to compulsorily acquire private land for public use, typically requiring fair compensation to the property owner. This power originates from the "Takings Clause" of the Fifth Amendment to the U.S. Constitution and must satisfy two constitutional requirements: "public use" and "just compensation." Historically, eminent domain has been used to build highways, parks, schools, and other public infrastructure. In the 2005 Supreme Court case Kelo v. City of New London, "public use" was broadly interpreted as "public purpose," allowing land seizure for economic development. Since then, over 40 states have passed legislation limiting this power. Nashville's "reverse application" of eminent domain — not to bring in a development project, but to protect public space from commercial encroachment — is itself an unusual policy signal in terms of legal strategy.

Why Data Center Expansion Sparks Community Controversy
The Potential Impact of Building a Data Center Next to a Zoo
Data centers are the physical foundation of modern AI and cloud computing. Training and inference for large models like ChatGPT, enterprise cloud services, and even everyday video streaming all depend on data centers distributed around the world to provide compute power. However, these facilities are not without cost.
A typical hyperscale data center — one with more than 5,000 servers and over 100,000 square feet of floor space — has staggering energy and resource demands. For training a GPT-4-class model, for example, a single training cluster may require tens of thousands of GPUs running simultaneously for months, with each high-end GPU (such as the NVIDIA H100) having a thermal design power of up to 700 watts. A 100-megawatt data center consumes roughly as much electricity per year as 80,000 American households combined. On the cooling front, traditional air-cooled data centers have a PUE (Power Usage Effectiveness — the ratio of total data center power to IT equipment power) of approximately 1.3–1.5, and evaporative cooling systems can consume millions of gallons of water per day. As AI workloads explode, single-campus power demands are leaping from the traditional 30–50 megawatts to 300 megawatts or even gigawatt-scale levels.
When such a facility is sited near environmentally sensitive areas like a zoo, its potential impacts become the focal point of community opposition:
- Noise may disrupt animals' normal behavior
- Competition with public recreational spaces for precious land resources
- Alteration of the surrounding area's environment and landscape
Why Local Governments Are Taking a Harder Stance
Nashville's choice to invoke eminent domain rather than simply relying on conventional zoning to block the project indicates a fundamental shift in local government attitudes toward data center projects. Zoning is the standard tool local governments use to manage land use, controlling development types by dividing cities into residential, commercial, industrial, and other functional zones. Amending zoning ordinances typically requires public hearings, planning commission review, and city council votes — a time-consuming process that developers can challenge in court. By contrast, once a government demonstrates that eminent domain serves a "public use," the landowner can only dispute the amount of compensation, not the seizure itself. Nashville's willingness to bear the fiscal cost (compensation fees) to ensure a definitive outcome reflects local decision-makers' belief that blocking this project carries sufficiently high public value.
The message is clear: not all tech investment will be unconditionally welcomed by local communities. When a project's negative externalities are deemed to outweigh its tax revenue and employment benefits, communities are fully capable of saying "no."
The Intensifying Conflict Between Compute Expansion and Community Interests
Anti-Data Center Sentiment Emerges Across the U.S.
Nashville is not an isolated case. As the AI arms race intensifies, from Virginia's "Data Center Alley" to Texas and Arizona, data center applications have surged and community opposition is growing in multiple regions.
Loudoun County in Northern Virginia, known as the "Data Center Capital of the World" and handling roughly 70% of global internet traffic exchange, offers a cautionary tale after more than two decades of rapid growth. Local residents have long complained about the hum of data center transformers, emissions from backup diesel generators, and community landscapes dominated by massive gray buildings. Between 2022 and 2023, Loudoun County and neighboring Prince William County successively passed stricter zoning restrictions, significantly increasing buffer distance requirements and noise standards for data center construction. Prince William County even rejected a large project near a Civil War battlefield site amid strong community opposition, demonstrating that cultural heritage preservation can also serve as a powerful reason to resist data center expansion.
Core concerns among residents include:
- Grid pressure: Large data centers' power demands may drive up local electricity prices or even affect supply reliability
- Water resource consumption: In arid regions, cooling water directly competes with residential water supplies
- Land and landscape: Massive server buildings transform community character while creating relatively few jobs
- Tax incentive controversies: Many tech companies receive enormous local tax breaks without necessarily returning commensurate public value
Notably, there is a "jobs paradox" in the data center industry: a multi-billion-dollar hyperscale data center may employ thousands of construction workers during the building phase, but typically requires only 50–100 full-time employees for equipment maintenance and security management during operations. A 2023 McKinsey study noted that data centers create only about 0.5–1 permanent jobs per megawatt of IT load — a ratio far lower than most other commercial real estate types. This leads many communities to find, when weighing pros and cons, that data centers' economic contributions are insufficient to offset the strain they place on the power grid, water resources, and community environment.
The Resource Allocation Battle for AI Infrastructure
From a broader perspective, this event reveals an often-overlooked dimension of AI infrastructure development: compute doesn't materialize from thin air — it is rooted in specific land, electricity, and water resources. When the tech industry expands rapidly in the name of "progress" and "innovation," it is often local communities that ultimately bear the physical costs.
Between 2024 and 2025, global tech giants' investments in data center infrastructure have grown explosively. Microsoft announced over $80 billion in spending on AI data center construction in fiscal year 2025, while Google, Amazon, and Meta each planned tens of billions of dollars in capital expenditure. According to industry analysts, the global data center market was valued at approximately $350 billion in 2024 and is expected to double by 2030. This unprecedented investment boom means hundreds of new data center projects will seek locations in the coming years, while suitable sites — needing to simultaneously meet requirements for adequate power supply, water availability, network connectivity, and reasonable land costs — are becoming increasingly scarce, further intensifying land competition between tech companies and local communities.
Nashville's choice reminds us that behind the AI boom lies a real resource allocation battle. Local governments, as guardians of public interest, are reassessing: what exactly does a data center bring to a community, and what does it take away?
Implications for the Tech Industry's Data Center Siting
For tech companies eager to expand their compute footprint, the Nashville incident is a wake-up call. Data center site selection is no longer merely a technical calculation of cost and network latency — it is a comprehensive game requiring full consideration of community relations, environmental impact, and political risk. The old model of ignoring local concerns and pushing projects through with tax incentives is facing mounting resistance.
In the future, the key to successfully landing data center projects may lie in whether companies can build genuine mutual trust with communities, specifically including:
- Making transparent and verifiable environmental commitments
- Providing reasonable resource-use compensation plans
- Making substantive contributions to local employment and the economy
- Conducting thorough community engagement during the site selection phase
When powerful public instruments like eminent domain begin being used against private tech projects, the entire industry needs to rethink its social contract.
Conclusion
This land dispute near the Nashville Zoo may be small in scale, but it is a highly representative microcosm of the AI era. It reminds us that on the road to pursuing compute and intelligence, technological progress and community well-being do not always move in the same direction. Finding the balance between data center construction needs and local public interests will be a long-term challenge that the tech industry and local governance must face together.
Related articles

Claude Code vs Codex: A Deep Comparison to Help You Choose the Right AI Coding Assistant
Deep comparison of Claude Code vs Codex: architecture differences, behavior patterns, and use cases. Based on SWE-RPG benchmark data, choose the right AI coding assistant for your team.

Meta's Alleged Addictive Design: A Full Breakdown of the Hook, Hold, Harvest, and Hide Strategy
Meta lawsuit reveals a four-step product design strategy: Hook, Hold, Harvest, Hide. A deep analysis of addictive design in the attention economy and its ethical implications for the AI era.

Running an AI Coding Agent on an Amiga 500: How 1987 Hardware Connects to Modern AI
A developer ran an AI coding agent on a 1987 Amiga 500 with a 7MHz CPU and 1MB RAM. Learn how client-server architecture enables vintage hardware to access modern LLMs.