Potential Impacts and Questions Communities Should Investigate
Working draft — Version 0.5
Last updated: July 23, 2026
Purpose
Large data centers can bring investment, tax revenue, construction work, and digital infrastructure, but they can also create substantial demands on electricity, water, land, public infrastructure, and surrounding communities. This document distinguishes between well-established impacts, project-dependent risks, emerging evidence, and unverified or overstated claims. The central question is not whether every data center creates every impact — it is whether a particular proposal has disclosed enough information for the public and decision-makers to evaluate its likely benefits, burdens, alternatives, and safeguards.
1. Electricity Demand
Potential impact: Large data centers can consume very large amounts of electricity and may significantly change local or regional electricity-demand forecasts.
What is supported: LBNL estimated data centers consumed ~4.4% of U.S. electricity in 2023, projected to rise to ~6.7–12% by 2028, with a later update estimating ~9.5–15.3% by 2030. A single hyperscale campus can request hundreds of megawatts of continuous power.
Verification sources:
energy.gov
LBNL 2024 report
LBNL 2025 update
DOE: Powering AI and Data Center Infrastructure
2. Grid Reliability and Infrastructure Strain
Potential impact: Rapidly adding a large, continuous electrical load can create generation, transmission, interconnection, and resource-adequacy challenges if sufficient infrastructure is not available before the facility begins operating. It is too broad to say data centers automatically cause “grid instability” — flexible demand, storage, redundant supply, and enforceable interconnection conditions can reduce risk.
Verification sources:
DOE: Powering AI and Data Center Infrastructure
LBNL: Speed to Power
3. Electricity Rates and Cost Shifting
Potential impact: Existing utility customers may face higher rates if required to pay part of the cost of infrastructure built primarily for a large data-center load. Higher residential bills are not inevitable — outcome depends on tariff design and enforceability.
Verification sources:
LBNL: Electricity Rate Designs for Large Loads
LBNL: Pricing/Service Agreements Brief
4. Fossil-Fuel Use and Greenhouse-Gas Emissions
Potential impact: Data-center electricity demand can increase greenhouse-gas emissions when additional electricity is supplied by fossil-fueled generation, and may contribute to prolonged operation of gas plants and pipelines. Claims should distinguish direct emissions, purchased-electricity emissions, marginal-generation emissions, and embodied emissions.
Verification sources:
LBNL 2024 report
GAO-25-107172
5. Backup Generators and Onsite Power
Potential impact: Diesel/gas engines and turbines used for backup, testing, or grid support can emit air pollutants and greenhouse gases. Diesel exhaust is associated with respiratory and cardiovascular harm and cancer risk. Note: “methane emissions from diesel generators” is misleading — diesel generators principally emit CO2, NOx, particulate matter, and CO, not methane.
Verification sources:
EPA: Clean Air Act Resources for Data Centers
EPA: Stationary Engines Fact Sheet
EPA: GHG Equivalencies Calculator
6. Air Pollution and Public Health
Potential impact: Air pollution from onsite engines, construction equipment, added generation, and traffic can contribute to respiratory and cardiovascular harms. A national modeling estimate is not proof of a local “cancer cluster” — local conclusions require facility-specific emissions, dispersion modeling, and epidemiological methods.
Note on the circulating “$1.5 billion” claim: This figure appears to refer to modeled national treatment costs from data-center-related air pollution in 2023, not a demonstrated local healthcare-price increase. It needs a complete primary-source review before a precise figure is used.
Verification sources:
EPA: Diesel Exhaust Impacts
EPA: Clean Air Act Resources for Data Centers
GAO-25-107172
7. Water Consumption
Potential impact: Some data centers can consume substantial quantities of water for cooling and indirectly through electricity generation. Demand varies significantly by cooling technology, climate, and electricity mix. Annual figures can conceal seasonal impacts during drought or peak demand.
Verification sources:
LBNL 2024 report
LBNL: Water Efficiency
LBNL: Water Use of Data Center Workloads
GAO-25-107172
8. Water Quality, Wastewater, and Chemical Management
Potential impact: Data centers may generate wastewater containing concentrated minerals or treatment chemicals; fuel, coolant, and battery chemicals can create spill risks. It is inaccurate to state that all data centers pollute water — many operate closed-loop or low-water systems.
Verification sources:
LBNL: Water Efficiency
CDC: Cooling Towers Module
9. Cooling Towers and Legionella
Potential impact: Facilities using cooling towers must manage Legionella growth risk through water-management and monitoring practices. This is not unique to data centers and doesn’t mean cooling towers routinely cause outbreaks.
Verification sources:
CDC: Cooling Towers Module
CDC: How Legionella Spreads
CDC: Water Management Programs
10. Stormwater, Erosion, and Flooding
Potential impact: Large campuses can replace vegetation and permeable soil with impervious surfaces, potentially increasing runoff, erosion, and downstream flood risk. More accurate to describe “potential increased runoff and flood risk” than to claim inherent flooding.
Verification sources:
EPA: Construction General Permit
EPA: Stormwater Discharges from Construction
11. Noise, Low-Frequency Sound, and Sleep Disturbance
Potential impact: Fans, chillers, cooling towers, transformers, and generators can produce continuous, tonal, and low-frequency noise causing annoyance, sleep disturbance, and reduced quality of life. Claims that inaudible infrasound universally causes significant disease are not established.
Verification sources:
WHO: Environmental Noise Guidelines
EPA: Noise Pollution Resources
12. Light Pollution
Potential impact: Security and industrial lighting can increase nighttime glare and skyglow, affecting residents and wildlife. Significance is site-specific and reducible through shielding, dimming, and setbacks.
Verification sources:
National Park Service: Light Pollution
DarkSky: Model Lighting Ordinances
13. Local Heat and Microclimate Effects — Emerging Evidence
Potential impact: Data centers convert most electricity into heat, potentially contributing to localized warming. A 2026 preprint reported ~2°C average increase near AI data centers, but this is emerging research requiring independent replication.
Correction to a circulating claim: The claim that data centers “raise temperatures by 9-16 degrees on average” is not supported.
Verification sources:
Marinoni et al. (arXiv preprint)
Molla et al. (ScienceDirect)
14. Land Use, Agriculture, Habitat, and Soil
Potential impact: Large campuses and supporting infrastructure can displace other land uses, compact soils, and fragment habitat. Claims of crop injury from ordinary operation require local evidence.
Verification sources:
USDA NRCS: Soil Data and Maps
USFWS IPaC
15. Construction Traffic, Road Safety, Dust, and Debris
Potential impact: Large construction projects can generate years of heavy-truck traffic, road wear, dust, and congestion.
Verification sources:
FHWA: Work Zone Safety
EPA: Stormwater Discharges from Construction
16. Employment and Local Economic Effects
Potential impact: Data centers can create substantial construction employment but often require fewer permanent onsite workers than similarly sized industrial uses. “Few permanent jobs” should be evaluated against acreage, subsidy, and investment — not presented without a denominator.
Verification sources:
Georgia Dept. of Audits: Data Center Tax Incentive Evaluation
Data Center Coalition: Reports and Publications
17. Tax Incentives, Public Funds, and Fiscal Risk
Potential impact: Tax exemptions, abatements, and other incentives can reduce public revenue or expose governments to financial risk. “Diverting public funds” is a political description, not a precise finding — a professional review should quantify recipient, value, and consequences of failure.
Verification sources:
Georgia Dept. of Audits: Data Center Tax Incentive Evaluation
Good Jobs First: Cloudy Data, Costly Deals
18. Property Values and Neighboring Land Uses
Potential impact: Noise, scale, traffic, and transmission infrastructure may affect nearby property desirability. Current evidence assessment: plausible and project-dependent; not established as a universal effect.
19. Depreciation, Stranded Assets, and Decommissioning
Potential impact: Data-center equipment can become obsolete quickly, and facilities may become underused if demand forecasts change. Accounting depreciation of equipment is not the same as depreciation of land or nearby property.
Verification source:
LBNL: Electricity Rate Designs for Large Loads
20. Transparency, Confidentiality, and Nondisclosure Agreements
Potential impact: Communities may be asked to approve projects without access to key information about ownership, load, water demand, or public costs. Security-sensitive details may need protection, but basic impact data and financial commitments should not be hidden merely because a project is commercially valuable.
21. Security, Cybersecurity, and Critical-Infrastructure Risk
Potential impact: Data centers are part of critical digital infrastructure and may host government, healthcare, financial, or communications systems. Cyberattacks, physical attacks, or prolonged outages can have consequences beyond the site. Calling every data center a “terrorist target” is speculative and unnecessarily alarming.
Verification sources:
CISA: Critical Infrastructure Security
GAO-23-105468
22. Surveillance and Privacy Around the Site
Potential impact: High-security facilities may use cameras, access-control systems, ALPR, or biometric systems affecting workers, visitors, and passersby. It is not supportable to describe all data centers as surveillance operations — the relevant concern is scope, retention, and sharing of security-system data.
23. The Data or Computing Uses Hosted at the Facility
Potential impact: Computing infrastructure can support beneficial, neutral, harmful, or unlawful activities. A data-center building is general-purpose infrastructure — unless customers/workloads are known, it is not supportable to claim a facility stores “problematic data” or powers a surveillance state. Claims connecting data centers categorically to serious criminal material are inflammatory and unsupported.
Verification sources:
GAO-25-107172
NIST: AI Risk Management Framework
24. Electromagnetic Fields and Wildlife — Insufficient Evidence for a General Claim
Circulating claim: Electromagnetic fields from large data centers disrupt animal behavior reliant on geomagnetic cues.
Current evidence assessment: Some animals use Earth’s magnetic field for navigation, but this does not establish that data-center EMF causes meaningful wildlife disruption. A data-center-specific claim requires measured field strength, exposure pathways, and relevant biological research.
Recommended wording: New electrical infrastructure should be evaluated for site-specific effects on habitat and wildlife. Claims that data-center electromagnetic fields broadly disrupt animal navigation remain unverified.
25. Claims That Should Not Be Repeated Without Better Evidence
The following statements are currently too broad, misleading, or unsupported to include as established facts:
– “Data centers cause cancer clusters.”
– “Healthcare costs near data centers rise significantly.”
– “Data centers raise temperatures by 9-16 degrees on average.”
– “Methane emissions are the principal problem with diesel generators.”
– “Data centers broadly disrupt animal navigation through electromagnetic fields.”
– “Every data center is a terrorist target.”
– “Data centers are the nexus of the surveillance state.”
– “All data centers pollute water.”
– “All data centers destabilize the grid.”
– “All data centers increase everyone’s utility bills.”
– “All data centers devalue surrounding property.”
– “Infrasound from data centers causes significant disease in people and animals.”
– “Data centers necessarily enrich billionaires at everyone else’s expense.”
26. Cross-Cutting Questions for Every Proposal
Ownership and scope: Who owns the land, facility, operating company, and equipment? What parent companies or subsidiaries are involved?
Actual resource demand: What are the initial, average, peak, and full-build electricity and water demands? Will actual use be disclosed after operation begins?
Supporting infrastructure: What power plants, transmission, substations, pipelines, roads, or water systems are required?
Public cost and risk: What tax benefits and public spending are involved? Who pays if the project is delayed or cancelled?
Environmental performance: What enforceable limits apply to emissions, water use, noise, and lighting? Who monitors compliance?
Community protection: How close is the project to homes, schools, farms, and sensitive habitat?
End of life: What financial assurance covers decommissioning and cleanup?
27. Suggested Minimum Public Disclosure Standard
Before final approval, a large data-center proposal should publicly disclose: legal and beneficial ownership; site acreage and buildout phases; initial/average/peak/max electricity and water demand; cooling technology; backup generation capacity, fuel, and emissions; required supporting infrastructure; expected permanent and construction jobs; tax incentives and public costs; cost-allocation protections for other ratepayers; environmental studies (air, water, noise, heat); emergency-response and decommissioning plans; financial assurance and clawback provisions; annual reporting of actual usage; and a public complaint-resolution process.
28. Evidence and Editorial Standards for Future Additions
When adding a claim to this document, prefer sources in this order: statutes, regulations, permits, and official records; government agencies and national laboratories; peer-reviewed research; university and nonprofit research with disclosed methodology; reputable investigative journalism; industry studies, clearly labeled as industry-funded; advocacy research, checked against primary sources; personal reports or social-media posts, used only as leads. Record the exact claim, geographic scope, time period, and important limitations. Avoid turning “can” into “does,” turning a maximum value into an average, or applying an AI-specific finding to every data center.
29. Intake Template for New Community Submissions
For each new item: original wording; reframed neutral, testable claim; evidence status (well established / project-dependent / emerging evidence / anecdotal / unsupported / needs further research); what appears verifiable; questions to investigate; documents needed; primary and secondary sources; important limitations.
30. Research Queue
Items requiring additional primary-source work: the underlying study behind the “$1.5 billion in treatment costs” figure; peer-reviewed evidence on local air-temperature effects; independent property-value studies; frequency-specific noise studies; comparative permanent employment per acre/megawatt/subsidy dollar; decommissioning cost examples; data-center-specific EMF/ecological research; documented Legionella incidents; traffic collision datasets; agricultural productivity evidence; best-practice security-camera privacy rules; hazardous-material incident data; model community-benefit agreements.
Working Conclusion (as of Version 0.5)
Data centers are not a single uniform category. The strongest community document will not claim that every facility causes every possible harm — it will establish that large data centers can create significant public consequences and that approval should depend on transparent, project-specific evidence, fair cost allocation, enforceable safeguards, independent verification, and continuing public reporting.
31. Transmission Corridors, Easements, and Displacement Risk
Potential impact: New transmission lines, substations, and access roads built to serve large loads can cross private property and, in some cases, lead to compulsory easements or acquisition. Electric-transmission siting is governed primarily by state law. It is too broad to say data centers themselves routinely “force people out of their homes” — the legal actor is generally a utility or transmission developer, and the line may serve multiple loads.
Verification sources:
FERC: Siting Interstate Transmission Facilities
FERC: Transmission Facilities Permit Process
32. Consumer-Electronics Prices and Component Competition — Plausible but Not Yet Generalizable
Potential impact: Rapid AI-infrastructure growth can increase demand for processors, memory, and networking equipment, potentially raising consumer-electronics prices where inputs are shared.
Current evidence assessment: Plausible and increasingly observable in particular component markets, but not established as a universal data-center impact.
Verification source:
Reuters: Intel, AMD Sign Long-Term Server CPU Deals
33. Community Dependence on Voluntary Corporate Benefits
Potential impact: In communities with limited tax bases, donations and voluntary benefits from a large operator can become financially or politically important, weakening independent oversight if essential programs depend on continued company goodwill.
Verification sources:
Brookings: Community Benefit Agreements
WRI: 7 Ways Data Centers Affect U.S. Communities
34. Allocation of Capital, Labor, Land, and Public Capacity
Potential impact: Very large investments may compete with other uses for capital, skilled labor, land, and public attention.
Current evidence assessment: The underlying scarcity and opportunity costs are verifiable. Whether a particular project constitutes “misallocation” is a policy judgment requiring comparison with realistic alternatives.
35. Public Sentiment, Opposition, Moratoria, and Project Delays
Potential impact: Growing concern about electricity prices, water, and environmental effects can produce organized opposition, moratoria, and project delays.
What is supported: Morning Consult (May 2026) found 69% of registered voters said AI data centers were very/somewhat responsible for rising electricity prices, and support for local construction bans rose to 42%. Reuters/Ipsos (June 11, 2026) found 57% opposed a data center in their community. Data Center Watch reported ~75 projects worth ~$130 billion blocked or delayed in Q1 2026.
Important qualifications: Data Center Watch is an opposition-tracking organization, not a government statistical agency. Projects may be delayed by grid constraints, financing, or permitting — not solely opposition.
Verification sources:
Morning Consult: Voters Blame AI Data Centers for Power Bills
Reuters/Ipsos: Americans Wary of AI-Driven Data Center Boom
Data Center Watch: Q1 2026 Report
Yahoo Finance: Growing Public Opposition to Data Centers
36. Additional Research Queue Items From This Submission
Obtain the Morgan Stanley research note on data-center opposition and verify its methodology; reconstruct Data Center Watch’s project-level list to test for duplicate counting; locate rigorous estimates of AI semiconductor demand’s effect on consumer prices; compare community-benefit agreements with enforceable obligations; collect documented transmission-driven displacement cases; develop a framework for measuring opportunity cost.
37. Cumulative Impacts of Supporting Infrastructure
Potential impact: The full impact of a large data-center project may extend far beyond the facility itself — new power plants, substations, transmission corridors, pipelines, water systems, and battery installations may each create additional burdens.
38. Transformer and Grid-Equipment Supply Constraints
Potential impact: Large projects can increase demand for transformers and grid equipment already in short supply, potentially delaying housing connections, renewable projects, and reliability upgrades elsewhere.
Verification source:
DOE: Keeping the Lights On in America
39. Wildfire, Smoke, and Emergency Power Risk
Potential impact: In wildfire-prone regions, new transmission corridors and backup-generation systems can create additional ignition, smoke, and evacuation concerns.
40. Fire Service and Hazardous-Materials Burdens
Potential impact: Large data centers may require specialized emergency response for high-voltage equipment, transformer oil, lithium-ion batteries, and chemical releases.
41. Competition for Emergency Fuel
Potential impact: Large facilities with substantial backup-generation capacity may compete for diesel fuel during regional emergencies, potentially affecting hospitals and fire departments.
42. Large-Load Grid Behavior and Sudden Load Loss
Potential impact: Very large electronically controlled loads can affect grid voltage and frequency stability. Sudden disconnection of a large load may create operational challenges.
Verification source:
DOE: Data Center EMT Models
43. Interconnection Queue Displacement
Potential impact: A large data-center request may reserve transmission or distribution capacity that could otherwise serve housing, manufacturers, or renewable-energy projects.
44. Water Infrastructure Cost and Stranded Capacity
Potential impact: A data center may require new wells, pipes, or treatment facilities. The public may inherit long-term maintenance or stranded-capacity costs.
45. Water Withdrawal, Consumption, and Discharge Definitions
Developers may report withdrawal, consumption, and discharge as though interchangeable. Withdrawal is water taken from a source; consumption is water not promptly returned; discharge is water returned to a sewer or water body; peak-day use is maximum short-term demand. Communities should require separate reporting for each measure.
46. Cooling-System Resilience
Potential impact: Cooling systems may perform differently during drought, extreme heat, or municipal restrictions.
47. Regulatory Capacity and Independent Oversight
Potential impact: Local governments may lack expertise to evaluate utility economics, acoustics, hydrology, and cybersecurity.
Good-practice principle: The developer may fund independent review, but the public agency — not the applicant — should select and direct the work.
48. Contractual Lock-In and Long-Term Flexibility
Potential impact: Public bodies may enter long-term agreements based on uncertain assumptions about technology, demand, and prices.
49. Corporate Structure, Subsidiaries, and Liability
Potential impact: Multiple subsidiaries and contractors may make responsibility for taxes, cleanup, and obligations unclear.
50. Public Records and Democratic Accountability
Potential impact: Key information may be withheld under trade-secret claims, NDAs, or utility confidentiality rules.
51. Political Influence and Regulatory Capture
Potential impact: Companies may fund lobbying, campaigns, and research institutions involved in public debate. Financial support does not prove corruption — the concern is undisclosed conflicts and unequal access.
52. Unequal Distribution of Benefits and Burdens
Potential impact: Tax benefits may be spread across a jurisdiction while noise, traffic, and water stress fall on a smaller number of nearby residents, including renters, Tribal communities, and low-income residents.
53. Revenue Volatility and Equipment Depreciation
Potential impact: Expensive computing equipment may create high initial taxable value but depreciate rapidly or qualify for exemptions.
54. Local Construction Inflation and Project Displacement
Potential impact: A multibillion-dollar project can raise costs or delay housing, schools, and other public projects by increasing demand for skilled labor and materials.
55. Temporary Workforce, Housing, and Service Pressures
Potential impact: Large construction workforces can increase demand for rentals, hotels, schools, and emergency services even where permanent jobs are few.
56. Insurance, Municipal Credit, and Debt Risk
Potential impact: Dependence on one dominant customer may affect utility debt, municipal credit, and insurance costs.
57. Embodied Environmental and Labor Impacts
Potential impact: The footprint includes semiconductor fabrication, mining, manufacturing, and global labor conditions — not only onsite electricity and water.
Verification source:
GAO-25-107172 (full PDF)
58. Refrigerant Use and Leakage
Cooling systems may use refrigerants with high global-warming potential. Communities should request the refrigerant inventory, leak-detection plan, and analysis of lower-impact alternatives.
59. Battery Lifecycle, Fire, and Disposal Risk
Battery systems can improve resilience but introduce mining impacts, fire/toxic-smoke hazards, and disposal liability.
60. Biodiversity and Habitat Fragmentation
Potential impact: Data centers and supporting infrastructure may affect wildlife through habitat loss, fencing, light, noise, and bird collisions.
61. Waste-Heat Recovery and Opportunity Cost
Most electricity consumed becomes heat. In some locations it may be reusable for district heating, greenhouses, or water heating. Communities should ask whether heat recovery was evaluated.
62. Cloud Concentration and Systemic Outage Risk
Potential impact: When essential services depend on a small number of hyperscale operators, one outage or cyberattack can disrupt government, hospitals, and businesses widely.
Verification source:
GAO: Cybersecurity Resources
63. Insider Threats and Supply-Chain Security
Risks may arise from compromised hardware, malicious insiders, and weak remote-management controls.
Verification source:
NIST: Securing AI Data Center Architecture
64. Data Sovereignty and Jurisdiction
Potential impact: The physical and legal location of data affects which governments may compel access and which privacy laws apply.
65. Priority Additions for Community Review
Categories deserving especially close attention: cumulative impacts of supporting infrastructure; emergency response and hazard risks; transformer shortages and grid displacement; regulatory capacity and enforceability; temporary workforce and construction-cost pressures.
66. Wetlands, Streams, and Aquatic Ecosystems
Potential impact: Data-center campuses and supporting infrastructure may fill, drain, or alter wetlands, streams, and aquatic habitat.
67. Tree Removal, Shade Loss, and Site Heat
Potential impact: Removing mature trees can reduce shade, increase surface temperatures, and intensify localized heat effects.
68. Construction Dust and Particulate Matter
Potential impact: Site clearing, grading, and heavy-truck traffic can generate dust and fine particulate pollution during construction.
69. Thermal Pollution From Water Discharges
Potential impact: Where heated water is discharged to surface waters, temperature changes may affect dissolved oxygen and aquatic life. Many data centers do not directly discharge heated cooling water to surface waters.
70. Groundwater Drawdown, Aquifer Depletion, and Streamflow Effects
Potential impact: Large groundwater withdrawals may lower water tables, reduce well yields, and affect connected streams.
71. Water Rights and Allocation Conflicts
Potential impact: A large project may intensify competition among municipal, agricultural, Tribal, and ecological water users, particularly in drought-prone basins.
72. Land-Value Inflation and Property-Tax Pressure
Potential impact: Large industrial development can increase nearby land prices, potentially raising tax burdens on existing residents and farmers.
73. Mental Health, Stress, and Neighborhood Industrialization
Potential impact: Continuous noise, lighting, traffic, and loss of rural character may contribute to stress and reduced quality of life. It is appropriate to recognize psychosocial burdens without claiming specific psychiatric diagnoses or universal medical effects.
74. Weak Community Consultation and Procedural Fairness
Potential impact: Communities may be consulted only after major siting, utility, and tax decisions have effectively been made.
75. State Preemption and Limits on Local Authority
Potential impact: State law may restrict local authority over zoning, utility regulation, and moratoria, limiting a community’s ability to impose stronger protections.
76. Concentration of Computing Power and Market Power
Potential impact: Ownership of advanced chips and large-scale compute may concentrate among a small number of corporations, potentially reducing competition.
77. Copyright and Training-Data Governance
Potential impact: AI systems may be trained on copyrighted or personal material, creating disputes over consent, compensation, and attribution.
78. Labor Conditions in Data Labeling and Content Moderation
Potential impact: AI systems may rely on workers who label data or moderate disturbing content under low-paid, psychologically harmful conditions.
79. Deepfakes, Misinformation, and Information Integrity
Potential impact: High-capacity AI infrastructure can support convincing synthetic media used for fraud, harassment, or misinformation.
80. Automation and Job Displacement
Potential impact: Expanded AI computing capacity may accelerate automation of clerical, creative, technical, and service work.
81. Massive Public Subsidies for Highly Automated Facilities
Potential impact: Governments may provide large subsidies to facilities that create relatively few permanent onsite jobs. This issue combines labor, environmental justice, and fiscal-policy accountability.
82. Critical Infrastructure Vulnerability
Potential impact: A facility supporting essential services may become a point of physical or operational vulnerability.
83. Cybersecurity Risk
Potential impact: Data centers may be targeted through ransomware, credential theft, and attacks on operational technology.
84. Manufacturing Emissions and Global Supply-Chain Footprint
Potential impact: Manufacturing servers, chips, and cooling systems creates emissions, water demand, and labor risks outside the host community.
85. Consolidated Additions From Community Review
Recurring themes: habitat loss and aquatic impacts; construction dust and generator emissions; battery/e-waste/embodied carbon; drought competition and water rights; transmission expansion and displacement; workforce housing and farmland loss; sleep, stress, and road-damage burdens; subsidies, secrecy, and environmental justice; computing concentration, hidden labor, and monopoly power.
86. Environmental Claims Used to Justify Experimental or Unproven Technology
Potential impact: A project may be presented as “green” or “clean” while introducing unfamiliar, insufficiently tested technologies. Environmental claims can be used to discourage scrutiny of local risks and alternatives.
Recommended wording: Environmental or clean-energy claims should not substitute for project-specific evidence. Communities should receive enough information to evaluate unfamiliar technologies and failure modes before irreversible decisions are made.
Verification sources:
EPA: Environmental Claims Guidance
DOE: Community Benefits Plan Guidance
87. Excessive or Irreversible Loss of Farmland
Potential impact: Large campuses and supporting infrastructure can permanently convert productive farmland to industrial use.
Verification source:
USDA NRCS: Farmland Protection Policy Act
88. Procedural Overload and Unequal Capacity to Review Technical Records
Potential impact: Residents may receive thousands of pages of technical filings while having only a short period to submit meaningful comments. The concern is procedural inequality — fragmented releases, late revisions, and inadequate technical assistance.
Verification sources:
EPA: Public Participation Guide
CEQ: A Citizen’s Guide to NEPA
89. Severe Stress, Entrapment, and Mental-Health Crisis During Siting Conflicts
Potential impact: Land acquisition, eminent-domain threats, and prolonged industrial siting disputes may contribute to severe stress, anxiety, and a sense of entrapment. In individual cases, people may report suicidal thoughts or other acute mental-health crises, which should be taken seriously and addressed immediately.
Important qualifications: Evidence that data-center siting disputes as a category cause suicidal ideation or suicide is not presently established. This topic should not be sensationalized.
Immediate safety note: When a person expresses suicidal intent or appears in immediate danger, the priority is direct support. In the United States, call or text 988 to reach the Suicide & Crisis Lifeline. Call emergency services for an immediate threat to life.
Verification sources:
CDC MMWR: Suicide Rates by Industry and Occupation
988 Suicide & Crisis Lifeline
90. Cumulative Pollution Burdens and Environmental Justice
Potential impact: Data centers may be proposed in communities already experiencing substantial pollution and industrial burdens. Adding another large source of emissions or water demand can increase total exposure even when the project independently complies with individual permit limits.
What is supported: EPA defines cumulative impacts as the combined effects of chemical and non-chemical stressors on health and quality of life. EPA and GAO have documented that historically marginalized communities often face disproportionate pollution exposure.
Verification sources:
EPA: Cumulative Impacts
EPA: Cumulative Impacts Research
EPA: Cumulative Impacts Research Recommendations
GAO-19-543: Environmental Justice
91. Research Queue Additions From These Submissions
Identify documented cases involving first-of-a-kind technologies and unexpected community impacts; develop criteria for distinguishing legitimate environmental benefits from greenwashing; quantify farmland conversion in major data-center regions; compare public comment periods across jurisdictions; locate peer-reviewed research on mental-health effects of prolonged siting disputes; develop ethical guidance for collecting testimony involving severe distress; identify siting patterns relative to existing industrial and demographic burdens.