The AI data center boom is the next environmental crisis and it’s already starting

Massive energy use, water drain, toxic pollution, and corporate subsidies: Why communities are fighting back, and why New Jersey must act now

A new industrial boom is spreading across the country, but instead of smokestacks and steel mills, it is massive AI data centers. These warehouse-sized facilities may power the digital economy, but they come with enormous and largely hidden environmental costs.

Artificial intelligence may be marketed as the future, but the infrastructure behind it is anything but clean. These facilities consume staggering amounts of electricity and water, rely on toxic chemicals, run fleets of diesel generators, and increasingly require new fossil-fuel power plants just to keep their servers running.

Across the country, communities are starting to realize that the AI boom is not just about technology; it is about land use, pollution, energy demand, and corporate subsidies. In many cases, the same communities already overburdened by pollution are now being targeted as locations for these massive facilities.

Instead of accelerating the transition to clean energy, the explosive growth of AI infrastructure threatens to lock us into decades of new fossil fuel generation, increased water withdrawals, and rising utility costs that ordinary ratepayers may ultimately be forced to cover.

Developers and tech companies are racing to build these facilities as fast as possible, often with little transparency and minimal environmental oversight. Meanwhile, policymakers are scrambling to understand the scale of the impacts.

New Jersey is already seeing the warning signs. With roughly 80 data centers already operating in the state and many more proposed, the Garden State is rapidly becoming a hub for AI computing infrastructure. At the same time, subsidies, zoning changes, and fast-track approvals are helping accelerate their expansion.

Without strong safeguards, the rush to build AI infrastructure could repeat the same mistakes we saw with warehouse sprawl and fossil-fuel expansion, sacrificing open space, overburdening communities, and weakening environmental protections in the name of economic growth.

Communities across the country are starting to push back.

They are right to do so.

Because the environmental damage from data centers is real, growing, and happening faster, without laws and rules to control it.

Massive energy demand and grid strain

AI data centers are among the most energy-intensive facilities ever built.

A single large AI-focused data center can consume as much electricity as 100,000 homes. The largest facilities, 600 megawatts or more, use enough electricity to power 350,000 homes or more.

This explosive demand is already placing enormous stress on regional power grids. Utilities are delaying the retirement of fossil-fuel power plants or proposing new gas generation to meet this demand. Some proposals even suggest reopening closed nuclear facilities like Three Mile Island.

Instead of accelerating the transition to clean energy, the rapid growth of data centers risks locking in new fossil fuel infrastructure for decades, undermining climate goals.

In many cases, developers are proposing “behind-the-meter” gas-fired power plants that would directly supply electricity to these facilities. That means building new fossil-fuel infrastructure specifically to power AI computing, moving us backward at a time when we should be rapidly reducing emissions.

Water consumption and resource depletion

The water demand from hyperscale data centers is staggering.

Large facilities can consume 3 to 7 million gallons of water every day for cooling systems. Just one large facility can use as much drinking water as 4,000 homes or an entire town of 10,000 people.

In many communities, this places tremendous pressure on local drinking water supplies and wastewater infrastructure.

Traditional cooling systems also produce wastewater discharges known as “blowdown.” These discharges can contain concentrated minerals, treatment chemicals, and metals that may affect rivers, lakes, and municipal treatment systems if not properly managed.

Cooling towers can also create chemical drift, a mist containing these chemicals that can travel a quarter mile or more and settle on homes, lawns, and nearby vegetation.

At a time when many regions are already facing drought and water scarcity, allowing facilities that consume millions of gallons of water every day raises serious sustainability concerns.

Toxic chemicals and pollution risks

Few people realize that modern data centers rely on a wide range of chemicals for cooling, maintenance, and fire suppression. If these substances leak, are improperly handled, or are discharged into wastewater systems, they can pose risks to both public health and the environment.

Cooling system chemicals

Water-based cooling systems frequently require chemical additives to control bacteria, algae, and corrosion. These chemicals are eventually discharged through blowdown water, and chemical drift from cooling towers can travel a quarter mile to over a mile, potentially damaging vegetation and affecting nearby communities.

Examples include:

Biocides: Chlorine dioxide, isothiazolinones, glutaraldehyde

These chemicals kill bacteria and algae, but can also cause skin irritation, respiratory problems, and aquatic toxicity.

Corrosion inhibitors: Phosphate, nitrites, molybdate

These chemicals can contribute to eutrophication, oxygen depletion in waterways, and ecosystem damage.

Heavy metals: Copper, zinc, chromium

These metals can leach from pipes and equipment and accumulate in ecosystems and human tissue.

Fungicides and Algaecides

Triazole fungicides such as propiconazole are classified by the Environmental Protection Agency as possible or probable carcinogens and endocrine disruptors. Algaecides containing copper or ammonia can cause health impacts ranging from skin irritation to chronic liver and kidney damage.

Refrigerants and “forever chemicals”

Many high-density cooling systems used for AI computing rely on specialized refrigerants and synthetic cooling fluids.

Some systems use PFAS chemicals, often called “forever chemicals,” because they do not break down in the environment. PFAS contamination has been linked to cancer, reproductive harm, immune system damage, and other serious health problems.

Other systems rely on hydrofluorocarbons, or HFCs, such as R-134a and R-410A. These are extremely potent greenhouse gases, and even small leaks during maintenance or disposal can significantly contribute to climate change.

Some large industrial cooling systems use ammonia, which is less damaging to the climate but is highly toxic and dangerous if released.

Cleaning and fire suppression chemicals

Data centers also use industrial solvents, glycol ethers, and other chemical cleaning agents to maintain sensitive equipment.

Fire suppression systems may contain chemicals such as FM-200 or Novec 1230. Older facilities may still contain halon-based systems, which damage the ozone layer.

Air pollution, noise, and local impacts

Beyond chemical risks, data centers generate several forms of local pollution.

Most facilities rely on large banks of diesel backup generators to ensure uninterrupted power. These generators emit nitrogen oxides, particulate pollution, and carbon monoxide, pollutants linked to asthma, heart disease, and premature death.

Facilities also generate constant noise from cooling fans and mechanical equipment, along with 24-hour lighting and security systems that disrupt nearby neighborhoods.

These facilities are often built close to residential areas, creating significant quality-of-life impacts for surrounding communities.

Loss of open space and farmland

As demand for large development sites grows, developers are increasingly targeting rural areas and farmland. The AI data centers are large and getting larger, 1 million square feet up to 4.5 million square feet, or Quaker Bridge Mall to the World Trade Center. The new super data centers are 10 million square feet plus.

Massive warehouse-scale data campuses are replacing forests, open space, and agricultural land. These data centers use 100-200 acres for small to larger centers, 500-1,000 acres. These developments fragment wildlife habitat, increase stormwater runoff, and permanently alter local land-use patterns.

Environmental justice concerns

Many proposed data center sites are located in low-income communities or communities of color that are already overburdened by pollution.

These communities often receive few economic benefits while facing increased pollution, higher electricity demand, and rising infrastructure costs.

Without strong oversight, regulations, and siting criteria, the data center boom risks becoming the next environmental justice crisis.

Massive communications towers across New Jersey

In order to move massive amounts of data between facilities and end users, developers are also proposing large communications towers across the region.

These towers use ultra-low-latency microwave technology to transmit data quickly, often so financial traders on Wall Street can gain milliseconds of advantage over competitors.

One example was the proposed 500-foot tower on Ramapo Mountain in the Ramapo Reservation, intended to shave microseconds off stock trades between financial centers. A data center serving Merrill Lynch already operates in Mahwah.

There are now dozens of similar towers proposed or being built.

These towers would overwhelm scenic vistas and create environmental damage, particularly for migratory birds. In the Highlands region, mountain ridges block these microwave transmissions, prompting developers to build taller towers to maintain clear transmission paths between financial trading centers and AI networks.

Rapid Growth in the Tri-State Region

New Jersey

New Jersey currently hosts roughly 80 data center facilities, concentrated in North Jersey locations such as Secaucus, Newark, and Clifton.

The state has also created roughly $250 million in subsidies and incentives supporting data center development.

Major proposed projects include:

• Nebius/DataOne facility in Vineland, a 300-megawatt complex that just became the largest in the state and includes on-site gas-fired power generation, creating pollution and misery in the community.

• CoreWeave Kenilworth campus, a proposed 250-megawatt AI computing center expected to be operational by 2027.

• PNJ-01 and PNJ-02 projects in Middlesex County, expected to add roughly 90 megawatts of capacity.

In Gloucester County, development activity has begun shifting from traditional warehouses toward potential AI data center projects.

Several proposals are currently in planning stages, including:

Logan Township: A new data center is expected to open in 2027.

Monroe Township (Williamstown): The Planning Board recently granted preliminary approval for a 1.6-million-square-foot warehouse complex on the former Under the Stars Farm along the Black Horse Pike. Although proposed as warehouses by Hexa Builders, residents and environmental advocates believe the project could ultimately become a massive AI data center due to its scale and infrastructure requirements.

West Deptford – Residents and local officials held a town hall in January 2026 to discuss concerns about a proposed data center and its impact on local resources. Local zoning changes are also fueling development pressure. In March 2025, Monroe Township Ordinance O:02-2025 amended the redevelopment plan to explicitly allow data centers as a permitted use for a 95-acre redevelopment site.

First win: Community pushback is already occurring. In New Brunswick, residents successfully forced the cancellation of a proposed data center project after raising environmental and community concerns.

New York

New York lawmakers are considering a three-year statewide moratorium on large new data center projects while the state studies their environmental and energy impacts.

Pennsylvania

Pennsylvania is also seeing massive proposals, including multibillion-dollar data center campuses and redevelopment projects at former industrial sites.

Many of these sites are located near the Delaware River, raising concerns about water withdrawals that could impact water supplies in New Jersey. Air pollution from new or reopened power plants could also cross state lines, undermining years of pollution reductions.

The scale of these projects is unprecedented and raises major questions about energy supply, environmental oversight, and regional impacts.

You cannot have AI with stupid laws and no regulations —legislative response is needed

States are beginning to recognize the scale of the challenge.

In New Jersey, lawmakers recently passed transparency legislation requiring data centers to publicly report electricity and water usage on a quarterly basis. This will help policymakers and the public understand the true scale of their resource consumption.

This should lead to stronger laws regulating and setting standards for data center development.

Additional legislation targets “large-load” electricity users, requiring them to pay for the infrastructure upgrades needed to serve their facilities. This “cost-causer pays” approach helps prevent utilities from shifting these costs onto residential customers.

Data centers should be required to pay the full cost of infrastructure upgrades needed to serve them.

Other proposed policies include requiring new facilities to operate on 100 percent new renewable energy so they do not drain existing renewable power supplies. Facilities should also be required to reduce energy use through efficiency and conservation.

Federal policy moving in the opposite direction

While states are becoming more cautious, federal policy is moving in the opposite direction.

Recent federal actions seek to fast-track permitting for data centers deemed critical digital infrastructure.

Some proposals include waiving environmental standards, allowing facilities to build their own fossil-fuel power plants, and weakening air pollution rules. There have even been discussions about reopening coal plants, including federal subsidies to keep aging plants operating.

Without strong environmental oversight, these efforts to accelerate construction could weaken environmental protections and bypass communities.

The bottom line — stop the data center free-for-all before it’s too late

Artificial intelligence may be transforming the global economy, but its physical infrastructure is creating very real environmental damage here on the ground.

Data centers consume massive amounts of electricity and water.

They rely on toxic chemicals and diesel generators.

They drive the need for new fossil-fuel power plants.

They threaten open space, farmland, and environmentally sensitive areas.

They are often built in overburdened communities that already suffer from pollution.

And too often they are supported by taxpayer subsidies and fast-tracked approvals.

We have seen this pattern before.

First, it was floating nuclear power plants.

Then, the toxic waste incinerators.

Then, the power plants and pipelines.

Then the warehouses.

Now it is AI data centers.

Each time we are told the same thing, that we must sacrifice environmental protections for economic growth, that the impacts will be small, and that the benefits will outweigh the costs.

Each time, pollution grows, infrastructure expands, and the public is left paying the price.

We cannot allow the AI boom to become the next environmental disaster.

New Jersey and other states must put strong safeguards in place now before the buildout gets out of control.

That means:

• Moratoriums where necessary, until the full impacts are understood, and new laws and regulations are in place

• No fast-tracking of data center permits

• No weakening of environmental review

• No subsidies or corporate tax breaks for billion-dollar tech companies

• No new fossil-fuel plants built to power AI

• No siting in environmental justice communities

• No development in the Highlands, Pinelands, environmentally sensitive farmland, forests, or open space

• Strict limits on water use and wastewater discharge

• Bans on PFAS and other toxic chemicals

• Full public disclosure of energy, water, and pollution impacts

• Requirements that new facilities run on new renewable energy, not power taken from the existing grid

• Impact fees so developers pay the real cost of infrastructure upgrades

• Strict pollution standards for data centers, quick phase-outs of diesel backup systems

• Closed-loop or zero-discharge cooling systems

• Limits on water withdrawals from stressed supplies

• Use of recycled or gray water for cooling

• Clear zoning and siting standards for data facilities

• Safer cleaning chemicals and pollution controls

• Location requirements avoiding environmentally sensitive areas and overburdened communities

• Preference for redevelopment of brownfields, landfills, and previously contaminated sites near existing infrastructure

• Green building standards and pollution reduction

• Public benefits such as open-space preservation to offset environmental damage

• An end to corporate welfare, subsidies, and tax breaks for these projects

We also need to have an open and transparent process to get there. Public input and involvement are critical in the legislative and regulatory processes. At the local level, there needs to be a public role in developing planning and zoning standards and siting criteria for these facilities, and analysis of infrastructure if it can handle the demand.

There needs to be full disclosure of campaign contributions from AI and data center companies, contractors, and consultants. There should be clear ethical standards and a pay-to-play ban for candidates, elected officials, and government officials. There are billions of dollars at stake, and insider deals could easily taint or corrupt the process.

Communities should not be forced to sacrifice their health, their water supply, their open space, or their electric bills so tech companies and hedge funds can move data a few milliseconds faster.

Artificial intelligence may live in the cloud.

But the pollution, the power plants, the water withdrawals, and the environmental damage happen right here.

If we do not act now, the AI data center boom will become the next generation’s environmental crisis, if it’s not already, and once these facilities are built, the damage will be locked in for decades.

The time to set limits is now.

The need for stronger safeguards

Given the enormous environmental footprint of these facilities, policymakers must adopt stronger safeguards before allowing uncontrolled expansion.

Key reforms should include:

• Mandatory public reporting of energy and water use

• Requirements for new facilities to operate on new renewable energy

• Strict pollution standards for backup generators and phase-outs of diesel systems

• Closed-loop or zero-discharge cooling systems

• Limits on water withdrawals from stressed supplies

• Strong environmental justice protections

• Comprehensive environmental impact reviews

• Moratoriums while environmental impacts are studied

• Clear zoning and siting standards for data facilities

• Use of recycled or gray water for cooling

• Bans on PFAS and other hazardous chemicals

• Safer cleaning chemicals and pollution controls

• Location requirements avoiding environmentally sensitive areas and overburdened communities

• Preference for redevelopment of brownfields, landfills, and previously contaminated sites near existing infrastructure

• Green building standards and pollution reduction requirements

• Impact fees to cover infrastructure upgrades and host-community costs

• Public benefits such as open-space preservation to offset environmental damage

• An end to corporate welfare, subsidies, and tax breaks for these projects

The path forward

Artificial intelligence may be transforming the global economy, but its physical infrastructure carries very real environmental costs.

Data centers consume massive amounts of electricity and water, rely on toxic chemicals and diesel generators, and increasingly threaten open space and vulnerable communities.

Without strong oversight and careful planning, the rush to build AI infrastructure could undermine climate goals, raise energy costs for consumers, and create a new generation of environmental problems.

Communities across the country are beginning to push back. Moratoriums, strict siting criteria, and stronger regulations are being debated at both the state and local level.

We are already seeing the damage caused by warehouse sprawl, facilities often built on speculation that can later become data centers, or even ICE detention facilities.

Policymakers must listen and act before the AI data center boom becomes the next major environmental crisis.