Dr. Rania Masri, NCEJN Co-director, gave a variation on this speech on January 20, 2026 for Democracy Out Loud
There is a new monster spreading like a virus throughout this country, and especially in the South: data centers and the Artificial Intelligence that they power.
What are data centers?
Data centers house large groups of networked computer servers and related infrastructure designed to process, store, and distribute vast amounts of digital data. They also are a cash crop for fossil fuel expansion. They demand massive amounts of energy, driving construction of new gas pipelines, coal plants, and nuclear facilities, and drain needed water sources, in the process.
Are we the primary users of these data centers?
No.
Isn’t it hypocritical that I’m giving this talk on Zoom?
No.
Data centers primarily serve “enterprise clients” like the military, surveillance, Big Oil including ExxonMobile, AI companies, cryptocurrency exchanges, and AI porn.
And more to the point: Although 63% of the world’s population use the internet worldwide, according to a statistic from 2022, only 38.5% of internet traffic is made up of humans. The remaining 61.5% are non-human, including search bots, scrapers, crypto, hacking tools, and other human impersonators, little pieces of code skittering across the web.
And, by this year, 90% of online content will be generated by AI!
So what is Artificial Intelligence?
Well, AI, or ‘traditional Artificial Intelligence,’ has been around since the 1950s. It focuses on performing a specific task intelligently. Traditional AI, often called Narrow or Weak AI, focuses on performing a specific task intelligently. It refers to systems designed to respond to a particular set of inputs.
Generative AI is something else entirely. It is a type of artificial intelligence that creates new content—such as text, images, music, code, or videos—by learning patterns from large amounts of existing data. Sounds good? Well, not necessarily.
Because Generative AI creates new content, it requires much more energy than traditional AI. Generative AI systems require data centers that are uniquely and horrifically resource-intensive, consuming vastly more water and energy to fuel the computing power and storage capacity they require.
Big Tech companies are rushing to gobble up the scarce resources that data centers use: land, infrastructure, power, and water.
Analysts predict that by 2030, AI will consume nearly as much energy as the entire country of Japan uses now—and only half of this power will be able to be generated from renewable energy sources.
So what is Generative AI used for?
Tristan Harris, co-founder and executive director of the Center for Humane Technology, warns us about Generative AI. He says: “Floods of deepfakes and AI-generated content overwhelm our information environment.” Remember: 90% of online content will be generated by AI this year. Conservatively, 62% of internet traffic is non-human.
Listening to Trisan Harris, I got scared. He is scared.
He says, “I used to be very skeptical when friends of mine in the AI community worried about sci-fi scenarios of AIs scheming, lying, or trying to escape. But unfortunately over the last few months, we’ve seen clear evidence of what was previously the realm of science fiction happening in real life. We’re now seeing frontier AI models lie and scheme to preserve themselves when they are told they will be shut down or replaced; we’re seeing AIs cheating when they think they will lose a game; and we’re seeing AI models unexpectedly attempting to modify their own code to extend their runtime in order to continue pursuing a goal. So we don’t just have a country of Nobel Prize geniuses in a data center—we have a million deceptive, power-seeking, and unstable geniuses. … We’re currently releasing the most powerful, inscrutable, uncontrollable, omni-use technology we’ve ever invented, one that’s already demonstrating the self-preservation, deception and escape behaviors we previously thought only existed in science fiction movies, and we’re deploying it faster than we’ve deployed any other technology in history, under the maximum incentives to cut corners on safety.”
In an interview, Harris spoke about “how AI could trigger a global collapse by 2027 if left unchecked, and how Ai will take 99% of jobs globally and collapse key industries by 2030.”
Already, AI-powered surveillance tools have been used to monitor migrants, including international students who are speaking up for Palestinian human rights and protesting the ongoing Genocide in Gaza.
Already, we are seeing a marriage between the ICE thugs and Big Tech.
Here’s one example: Palantir.
Palantir was initially funded in 2022 by the US Central Intelligence Agency’s venture-capital arm to provide “data-analytics software to U.S. Immigration and Customs Enforcement (ICE)], the FBI, the Department of Defense, and a host of foreign-intelligence agencies.”
Palantir is involved in everything from government surveillance and predictive policing to gathering battlefield intelligence and monitoring nuclear safeguards. It is ensconced in the US government and its armed forces. It is everywhere the US is supporting wars, including against Russia, Palestine and Iran.
Palantir has also contributed to Israel’s genocidal war machine in Gaza. In January 2024, the company boasted about its “strategic partnership” with the Israeli Ministry of Defense to supply the country with technology for its military. In her report “[f]rom economy of occupation to economy of genocide,” Francesca Albanese, the UN special rapporteur on the situation of human rights in the Palestinian territory occupied since 1967, cited Palantir’s strategic partnership with Israel and concluded that there are reasonable grounds to believe that Palantir’s AI platform has been used in Israel’s “unlawful use of force,” causing disproportionate loss of civilian life in Gaza.
And ICE has partnered with Palantir Technologies to use artificial intelligence and data mining to identify, track, and kidnap suspected noncitizens – and citizens – and murder people on the streets or in detention private camps.. Yes – kidnap. Up to 1,000 individuals are kidnapped a day throughout the US today.
And, in NC, the Durham police department is looking into partnership with Peregrine’s CEO and founder, who previously worked for Palantir!
In a nut shell: AI is not a neutral or net good technology.AI is used to increase unconscionable abuses: exacerbating housing discrimination of Black communities, increasing scams targeting vulnerable elders, heightening state and corporate surveillance, monitoring workers’ every move, and support ICE thugs here and genocide abroad.
And – Generative AI has already caused a loss of jobs – and will cause much more unemployment. The most vulnerable to these threats in the US are those already marginalized: Experts say AI will disproportionately impact Black and brown workers through displacement.
In North Carolina, there are more than 100 data centers in operation already, seven proposed (as far as we know), and more are coming!
So – How do these data centers impact our communities?
Energy
Data centers use massive energy that destabilizes our electricity grid – increasing risk of our homes catching on fire and blackouts and brownouts that cause our food and insulin to spoil in the fridge – and increasing the price of our monthly electricity bills. US consumers will pay billions of dollars to build the new power plants and infrastructure needed to serve Big Tech, as data center energy use is expected to triple in the next five years. Electricity bills are projected to rise by 8% nationally and as much as 25% in some regions by 2030.
Data centers are projected to consume about 6.7% to 12% of US electricity in 2028, up from 4.4% in 2023.
The Union of Concerned Scientists reports that nearly $1 trillion in electricity costs will be attributable to data centers over the next 25 years.
Duke Energy submitted a request on November 20th to raise rates for both its NC utilities. It is asking for an annual increase of $1 billion for Duke Energy Carolinas ($727 billion in 2027, $275 in 2028), a 15% increase over current revenues, and $729 million for Duke Energy Progress ($528 million in 2027, $200 million in 2028), a 15.1% increase. It says these investments are needed in part to “support rapid growth from increasing population, advanced manufacturing and data centers.” Duke is proposing: Raising customer’s bill And increasing the portion going to shareholders
Meanwhile, there is an increased expansion into fossil fuels, natural gas, and nuclear power. Duke Energy recently submitted an early site permit application to the U.S. Nuclear Regulatory Commission for a potential nuclear power plant in Stokes County, just north of Belews Creek.
And there is more.
Water
A single data center can use up to 5 million gallons of water each day – equivalent to the daily use of a 50,000 person town. 5 million gallons of pristine water every day.
Data centers require fresh drinking water—other sources corrode equipment or breed bacteria—and the water cannot be recycled. Many centers are permitted to use several million gallons per day—more water than 49,000 Americans typically use.
And only half of operators track their water use at all.
Meanwhile, 40% of data centers have been sited in areas of high or extremely high water stress.
And, in NC, A new report from the North Carolina Drought Management Advisory Council shows 92 of 100 counties are experiencing drought conditions, with two eastern NC counties experiencing severe drought – Wilson and Edgecombe counties! Edgecombe County, one of two NC counties in severe drought, is where Energy Storage Solutions is proposing a $19 billion, 900 mega-watt data center. A twin project is also planned for Fayetteville.
As if that wasn’t bad enough …
There’s PFAS. Yes.
PFAS.
Data centers “contain rows and rows of computer servers that use “forever chemicals.”
As revealed in a Guardian article “data centers increase PFAS pollution directly and indirectly. The chemicals are needed in the centers’ operations – such as its cooling equipment – which almost certainly leads to some on-site pollution. Meanwhile, Pfas used in the equipment housed in the centers must be disposed of, which is difficult because the chemicals cannot be fully destroyed. Plus – a large quantity of Pfas are used to produce the semiconductors housed in datacenters, which will increase pollution around supporting manufacturing plants. … Currently, companies are not required to record the amount of PFAS chemicals they are using or discharging. The chemicals are added to the cooling process, and then discharged out into the community the data center is housed in.”
Another set of chemicals of concern in data centers is called fluorinated gases or “f-gases.” F-gases are potent greenhouse gases. For example, perfluorohexane, which is sometimes used in these cooling systems, is 10,000 times more potent than carbon dioxide. F-gas producer Chemours is using the boom in AI and datacenters as justification for increasing production at its Parkersburg, West Virginia, and Fayetteville, North Carolina, plants.”
And there’s more.
Noise
Data centers also are noisy! Data centers run 24/y, and their cooling fans run 24/7 – harming data center staff, nearby communities, and local wildlife. Data centers, typically, produce between 85 – 95 dBA, well above the World Health Organization guidelines on human noise exposure. Data center noise can trigger stress, sleep loss, anxiety, heart risks, and in severe cases, tinnitus or hearing loss.
In Edgecombe County, the county commissioner claims that the noise level for that data center proposal will not exceed 60 dBA. But continuous 60 dBA can still be problematic. It can cause: Fragment sleep cycles. Increase nighttime awakenings. Reduce deep and REM sleep. Worsen mental fatigue. Increase next-day fatigue and cognitive impairment. Even if the noise does not wake you fully, it activates stress responses in the brain.
And there is more.
Land
These data centers can take up hundreds of hundreds of acres of land. In Stokes County, recently, the county commissioners approved a plan for a 1,800 acre data center.
1,800 acres. 2.8 square miles. (7.2 square kilometers). Visualize football fields laid edge-to-edge:40 fields long. 34 fields wide. 1,360 football fields total.
How big is 1,800 acres.
You start walking: After 10 minutes, you’re still very much inside it. After 30 minutes, you’re maybe halfway across. After an hour, you finally reach the other side — and that’s walking straight, no stops.
1,800 acres. Is the size of a small town. An entire small town.
What else could that land provide?
Affordable housing.
Agriculture. Family farms.
Or farmland. “New Hill Digital Campus” wants to take 189 acres of farmland outside the town of Apex for a 300 MW data center.
Or Forests. Or what the US Department of Defense (now, more appropriately called the US Department of War) calls ‘undeveloped land. The DOW is now seeking private AI data center developments on military bases, including on 734 acres in Fort Bragg (Cumberland and Harnett), an area that is primarily mixed pine and hardwood forests.
As Greenpeace wrote in a recent statement: “The problem with data centers is not just about climate impacts, or energy, or water, or the dirty supply chain. It’s also about the massive economic burdens local residents are shouldering because of the buildout. It’s about land being taken from communities to serve corporate interests and billionaires. It’s about the refusal to respect residents through transparency and public participation. We know that we, the people, deserve better. What are you willing to sacrifice to support big tech’s buildout?”
Once again, it doesn’t stop here.
There is more.
These ugly eyesores taking over massive land space and expanding fossil fuel consumption and powering surveillance and predictive policing and more – also cause significant health impacts.
Health
Increases in cancer and asthma are expected. By 2030, air pollution from data centers alone could cause an additional 600,000 asthma cases annually in the US; An estimated 1,300 premature deaths.
Annual public health costs from data-center-related energy use could reach $5.7B–$9.2B.
Already, Big Tech’s existing data build-out has generated $5.4B in health costs.
And we’re not talking about the public health impacts of contaminated water, or decreased water availability, nor the public health impacts from noise pollution, or the significant impact on public health from unemployment and reduced financial stability. We’re just talking about air pollution.
And, yes, there is more.
The material used in building data centers also impacts other industries. For example: Data centers will consume 70 percent of memory chips made in 2026 The shortages are driven by explosive AI demand. Up to 70 percent of the memory [chips] produced worldwide in 2026 will be consumed by data centers. The exponential rise in memory is all but guaranteed to hit the automotive sector, TVs, and consumer electronics, among many others.”
Meanwhile, democracy is further threatened. What is democracy without information?
As recently explained, “the AI infrastructure boom is quietly rewiring how land is bought, valued, and governed — often before the public has its say. In many cases, local officials and residents don’t even know who the buyer is until the deed is recorded. Developers are scouring the map for places where electricity is available now — or can be made available faster than somewhere else — where water rights are secure, and where fiber can be brought online without years of delay. … Today, the most important factor in site selection is whether or not the grid can deliver massive, continuous electricity demand sans interruption … demand from hyperscale customers can materialize quickly once a site is selected. The result is a geography shaped by electrons rather than people, and land value pegged to megawatts rather than population growth. … Land purchases are frequently conducted through shell companies and non-disclosure agreements that function to obscure the ultimate buyer. By the time public debate catches up, the fundamental decisions have already been made—about land use, about infrastructure, about long-term resource demand.”
For context, as the article explains, this is not new. “In historical context, however, the pattern is much more familiar and of a piece with the past than you might guess. An often-overlooked principle of American capitalism is its tendency toward monopolistic forms of production.” But I digress.
All this is presented in the media, when these so-called challenges are even presented in the media, as ‘environment vs jobs.’ A completely false binary, once again.
But what about jobs? Is the promise of jobs true? Shall we sacrifice long-term health for short-term employment?
Let’s look at the numbers. Data centers create 100 times few jobs than other types of economic development, ranked by the amount of energy used. Few permanent jobs are actually created. The common claims of robust job creation from the ballooning data center industry are grossly inflated! While the industry consumed more than 4% of all electricity usage in 2024, they provided only 0.01% of all employment across the country. In Virginia, the investment was nearly 100 times greater than was what required to create a job outside of that industry!
Still, decision makers claim that the county (or city) can make money, needed revenue, through increased property taxes. For the three Stokes County Board of Commissioners who voted Jan. 12 to approve the two rezoning requests necessary as a first step forward [for Project Delta], the potential for a community wide revenue boost took precedence over the appeals of more than 350 constituents at the public hearing.
According to the article published in the Greensboro News & Record, “A single PowerPoint display crystalized the clash between economic and environmental interests over a controversial proposed multi-billion-dollar data center in Stokes County. It simply stated $7,833 — $40 million. The first number represents the current property tax revenue from an idyllic, historic and undeveloped 1,844-acre tract near Walnut Cove.The second is a property tax revenue estimate, from developer Engineered Land Solutions chief executive Drew Nations, that a 24/7 data center would generate annually at full buildout spread out over 1,000 acres.”
So, what about money to the county?
NC provides sales and usage tax exemptions for data centers. These are moneys taken from our communities; Public schools are funded primarily from state sales tax and local property taxes! 38 states provide sales tax exemptions for data centers – including North Carolina. These exemptions lead to significant losses in tax revenue. Ten states lost out on more than $100 million per year. Despite substantial data center growth, NC fails to disclose the full cost of these exemptions. No other form of state spending is so out of control.
So how much money is lost to the County from the millions or billions given away to data centers by the State?
And does the county always have increased revenue?
Google’s data center in Lenoir paid only $5 million in county taxes over 10 years but received $73 million back in tax rebates—14 times what the county received. The city also abated $51 million in property taxes over eight years (2017–2024).
It is quite clear: it makes no sense to support a data center, unless you are one of the billionaires benefiting from the data center. It makes no financial or economic sense. It harms our communities’ health, our land and water and air. It harms what remains of our democracy, while powering militaries and thugs.
Are we fighting a losing battle here?
No.
Let’s remember that there has been a 125% surge in data center opposition throughout the country. In the four months of 2025 – from March to June – an estimated $98 billion in projects were blocked or delayed, more than the total of all previous quarters since 2023, when $64 billion in projects had been blocked!
And small towns in NC are hearing the call.
Town leaders in Canton (Haywood County) have taken the first steps toward a moratorium or temporary ban on data centers and crypto mining. The town board passed the motion to set the public hearing for the moratorium on data center.
And Gates County approved a 12 month data center moratorium!
Given that more data center projects are likely to be introduced in NC in 2026, with the growing opposition to data centers, and the increased organization, hopefully more moratoriums will be placed in their way!
Even in SC – Energy legislation filed recently in South Carolina would bar future sales tax breaks for data centers, while keeping them in place for those that inked deals by May of last year as long as they meet paperwork requirements. All three of the sponsors are Republicans.
So, let’s organize, let’s build our strategy, let’s fight back, and let’s build a tsunami of political power that can – finally – put the people who represent public interest into power.
Are you working to stop a data center and/or AI? If you’re not yet a member of the NC Data Center Network, please reach out to Rania Masri.
For more reports on data centers and AI, please request access to this shared drive.
