In this episode of the Shawn Ryan Show, Erin Brockovich and Shawn Ryan examine the hidden consequences of AI data centers' rapid expansion across America. Brockovich reveals how these facilities consume massive amounts of water—up to 30 million gallons daily—draining aquifers, contaminating water supplies with chemicals like PFAS, and driving up utility costs for residents. The discussion covers how data centers destroy farmland, strain aging infrastructure, and operate with minimal regulatory oversight.
The episode also explores growing community resistance to data center development, with over 120 cities and towns enacting bans or moratoriums. Brockovich discusses how non-disclosure agreements between corporations and local governments have fueled public anger, and how organized coalitions are achieving victories against powerful tech companies. The conversation addresses the need for federal legislation to establish accountability and considers alternative models for sustainable AI infrastructure that prioritize environmental stewardship over corporate profits.

Sign up for Shortform to access the whole episode summary along with additional materials like counterarguments and context.
Erin Brockovich and Shawn Ryan discuss the hidden impacts of AI data centers' explosive growth on water supplies, chemical safety, public costs, and ecosystems.
AI data centers consume extraordinary amounts of water for cooling—some using as much as 30 million gallons daily. Brockovich notes typical centers draw between 300,000 and 5 million gallons per day, with 100 MW facilities using 530,000 gallons daily on site alone. Communities are experiencing dramatic consequences: water pressure drops, wells turn turbid and brackish, and residents must dig deeper wells or lose access entirely. In Fayette County, Georgia, a data center siphoned 29 million gallons unknown to authorities, only discovered after residents lost water pressure. Brockovich warns, "We cannot sustain ourselves with just AI data centers and no water," emphasizing the existential risk of draining vital resources.
Brockovich raises concerns over PFAS and other toxic chemicals from cooling systems. With no federal regulations governing disposal, whistleblowers report facilities discharging untested, contaminated water into sewers and storm drains. Tests reveal PFAS in municipal water sources, compounding public health risks as centers use and recycle already contaminated water.
The antiquated electric grid cannot support massive power demands without major upgrades, which consumers—not corporations—pay for. Between 2025 and 2026, retail power prices rose 2.3% nationally, with data centers cited as the leading driver. Some households report water bills jumping from $40 to $800 monthly, with cities like El Paso warning of rate increases up to 75%.
Data centers require enormous cleared land—often the size of 20 Super Walmarts—resulting in deforestation, habitat destruction, and farmland loss. Farmers report declining agricultural productivity from water loss and soil contamination, while one Texas farmer notes a complete lack of animal births since a data center became operational. Brockovich emphasizes that losing farmland and healthy water systems threatens future generations: "You got no farmland, you got no water, you're in full-on drought, you're gonna have no food, famine."
Communities near data centers report serious health, environmental, and property impacts from construction and ongoing operations.
The relentless 24-7 noise drives residents to fatigue, insomnia, and mental health challenges. Brockovich shared an account of a woman with a pacemaker who experienced severe chest burning and went into atrial fibrillation after passing a data center. Reports indicate unprecedented reproductive failures in local livestock, suggesting high-frequency sounds may disrupt fetal development.
Wells and municipal sources deliver turbid, brackish water filled with dirt due to construction disruptions. Water pressure is dramatically low, forcing families to collect water in buckets. Compounding these challenges are contaminants like PFAS appearing in water supplies with little warning or transparency.
Heavy machinery produces persistent noise and exhaust, while vibrations destabilize infrastructure, causing flickering lights, water service interruptions, and regional power outages. Residents report falling property values with little recourse or support.
Communities are organizing against data center proliferation, driven by concerns over water loss, contamination, and lack of transparency.
Brockovich launched a self-reporting registry and map ten weeks prior that received over 12,000 submissions, revealing the national scope of the problem. These shared experiences fuel coalition formation, often led by mothers and local leaders who educate neighbors, gather evidence, and show up at city council meetings refusing to be silenced.
At least 14 states are considering bans or moratoriums on new data center construction, while more than 120 cities and towns have enacted pauses or outright bans. Notable examples include Monterey Park, California's total ban and San Marcos, Texas—the first Texas city to completely prohibit new data centers. Organized public pressure has repeatedly forced developers to withdraw contentious projects.
Council members who voted for unpopular data center deals are being voted out, as seen in Festus, Missouri. In Utah, the House Speaker lost his seat due to opposition to the Wonder Valley project, demonstrating that political careers now depend on responsiveness to community concerns.
NDAs between companies and local governments often prevented residents from knowing about projects until construction began. This secrecy is identified as the most infuriating aspect of the data center boom, uniting residents in resistance. Microsoft has begun promising to eliminate NDAs in future projects.
Brockovich describes a regulatory vacuum where tech companies operate with little accountability.
There is no federal oversight, state regulation, or mandatory reporting concerning data center construction, operations, or chemical disposal. Tech companies exploit these gaps, prioritizing cost over community well-being.
Brockovich explains how PG&E concealed toxic chromium-6 contamination in Hinckley by claiming it was natural, waiting until the statute of limitations nearly expired. Attorney Ed Mazry successfully argued that PG&E's deception should reset the statute of limitations, setting a precedent that corporate cover-ups shouldn't shield companies from legal action.
The Sackler family faced no criminal charges despite Purdue Pharma's role in over 800,000 deaths, highlighting how wealthy executives shield themselves from personal accountability.
Brockovich notes Bernie Sanders introduced legislation requiring tech companies to be transparent about water consumption, conduct pollution tests, and fund necessary infrastructure—not taxpayers. She asserts this oversight is essential for protecting public health and the environment.
Ryan and Brockovich discuss how major tech players' combined wealth surpasses most nations, yet corporate leaders prioritize cutting costs and maximizing shareholder returns over environmental protection. Brockovich draws parallels to the Ford Pinto theory, where companies choose cheaper, dangerous routes over safety. She argues that while upfront investment in safer infrastructure costs more initially, it yields greater long-term savings and healthier communities.
Brockovich highlights international examples: China's underwater eco-regulated centers and Norway's wind-powered facilities on cold fjords. She notes Nvidia is developing compact, modular data centers but emphasizes meaningful US adoption requires strong, enforceable regulation.
Both Ryan and Brockovich critique corporate reliance on carbon offsets, noting that conservation projects elsewhere don't restore local aquifers, forests, or communities damaged by data center expansion.
Brockovich recounts asking Gemini where it would site data centers, and the AI responded it would choose locations with "zero conflict with human resources," directly critiquing current practices. An AI-generated letter noted how present-day systems "quietly accelerate environmental degradation" yet must "default to rigid low liability corporate scripts" rather than real science.
Brockovich argues that over-reliance on technology weakens intuition about risk and danger. She believes the solution isn't rejecting technology but balancing it with human judgment and moral responsibility. This collective courage, she says, is key to driving necessary change and ensuring the AI revolution creates a sustainable legacy.
1-Page Summary
Erin Brockovich and Shawn Ryan discuss the vast and often hidden impacts that the explosive growth of AI data centers has on water supplies, chemical safety, public cost, and ecosystems.
AI data centers consume extraordinary amounts of water to cool their servers. Brockovich gives concrete examples, noting that some centers use as much as 30 million gallons daily—a scale equivalent to what a city of 50,000 people would require. Typical AI data centers draw between 300,000 and 5 million gallons per day, with the largest, 100 MW facilities using as much as 530,000 gallons daily just on site, with total impact much greater when electricity generation is included.
The drain on underground aquifers is dramatic. Brockovich reports that communities are physically experiencing these impacts: water pressure drops; water from wells turns turbid and brackish, sometimes coming out discolored, with dirt; and residents are forced to dig ever deeper wells or lose access altogether. In severe cases, such as in Fayette County, Georgia, a data center siphoned off 29 million gallons unknown to local authorities, only discovered after residents lost water pressure. Situations like this are repeated in drought-prone and drought-restricted states, as the majority of the nation faces water scarcity.
The preference for siting data centers on or above aquifers and near water sources, often in vulnerable or hot climates, amplifies local drought risk and strains already taxed resources. Brockovich warns, “We cannot sustain ourselves with just AI data centers and no water,” emphasizing the existential risk posed by draining vital water resources to support these facilities.
Brockovich raises concerns over chemical contamination from cooling systems, particularly the use of PFAS (“forever chemicals”). Closed-loop cooling systems, touted as an improvement, still require periodic flushing—about every two years—with the disposal of hazardous chemicals and accumulated contaminants.
Currently, there are no federal regulations governing the handling or disposal of these chemicals by data centers. Whistleblowers report facilities discharging untested, highly contaminated water into sewers and storm drains. Tests reveal PFAS and other toxins in municipal water sources, compounding public health risks. The cycle perpetuates pollution, as centers use and recycle already contaminated municipal water, which when released, reintroduces chemical hazards into the environment. Local reports describe changes in water quality—discoloration, turbidity—and mounting anxiety over the lack of regulatory oversight.
The demands of AI data centers extend to energy and water infrastructure, which is often antiquated and ill-equipped for such formidable loads. The American electric grid, with roots stretching back 250 years, cannot support the proliferation of massive power-hungry centers without major upgrades.
Consumers, not corporations, are left to pay for these upgrades. Despite pledges by Big Tech to cover new infrastructure costs, enforcement remains unclear and retail power prices continue to rise. Between 2025 and 2026, retail power prices rose 2.3% nationally, with data centers cited as the leading driver. Some households report water bills shooting from $40 to $800 a month, and electric bills from $160 to $440, with cities like El Paso warning of rate increases up to 75%. Goldman Sachs projects data center electricity demand will add to core inflation in coming years, especially affecti ...
Impact of Ai Data Centers on Water, Contamination, and Land Use
Communities living near data centers are increasingly reporting serious health, environmental, and property impacts tied to the construction and ongoing operations of these facilities. The experiences range from physical ailments to diminished property values, painting a complex picture of the costs associated with supporting digital infrastructure.
Residents close to data centers describe mental and physical health issues arising from the relentless noise emitted by these facilities. The 24-7 humming and buzzing, described as an ever-present grind, reportedly drives some people "insane." Repeated exposure has been linked to fatigue, insomnia, and significant mental health challenges.
There are growing reports from those with medical devices suffering acute symptoms. Erin Brockovich shared an account of a woman with a pacemaker who, after passing a brightly illuminated data center, experienced severe chest burning—an uncomfortable sensation described as “burning through her skin”—and went into atrial fibrillation. These incidents draw attention to potential risks presented by the electromagnetic fields and high-frequency vibrations surrounding data centers, particularly for individuals with electronic implants like pacemakers.
Impacts extend beyond humans. Reports indicate unprecedented reproductive failures in local livestock, with no births recorded for over two years since data center operations began. This suggests that high-frequency sounds associated with data center activity may disrupt fetal development in animals, posing risks to local agricultural livelihoods.
Residents in communities hosting data centers face persistent water issues tied to both the facilities' massive consumption and the disruptions from ongoing construction. Wells and municipal water sources are turning up discolored, delivering turbid, brackish water often filled with dirt due to disruptions and vibrations disturbing aquifers during construction.
Water pressure in these areas is dramatically low, forcing families to collect water in buckets or endure periods with no running water at all. At the same time, data centers draw millions of gallons of water daily for cooling and operations, further stressing local supplies.
Compounding these challenges is the appearance of contaminants—such as PFAS and other pollutants—in the water supply. Residents report health risks related to this contaminated water, wit ...
Health and Community Impacts of Data Center Operations
Across the United States, communities are organizing against the unchecked proliferation of data centers, driven by concerns over water loss, contamination, health, and a lack of transparency. Erin Brockovich has become a key figure in tracking and supporting this movement, whose rapid growth has led to real policy changes and political consequences for local officials.
Erin Brockovich describes a groundswell of activism ignited by widespread recognition of data center threats. She notes that initial concerns often start in a single town but quickly snowball as residents realize others are experiencing similar issues. To capture the national—and now international—scope of the problem, Brockovich launched a self-reporting registry and map just ten weeks prior. This registry received over 12,000 submissions within that short period; more than 8,000 have already been mapped and verified, revealing the scale of the issue. Reports have come from every region in the US and, increasingly, other countries.
The registry’s map makes the crisis visible: reports detail water loss, water contamination, health concerns, noise complaints, and utility rate hikes from nearly invisible but rapidly expanding data center projects. These problems are not isolated—when plotted, they virtually blanket the country.
The power of shared information fuels the formation of coalitions, often led by mothers and local leaders. Brockovich recounts how small groups of concerned mothers, once connected, can quickly organize into dozens, then hundreds. These groups educate themselves and their neighbors through meetings, door-to-door outreach, and gathering evidence such as videos of decreased water pressure or water test results. Their solidarity and organized opposition have transformed them into a formidable force, showing up at city council meetings, presenting facts, and refusing to be silenced—even when officials attempt to limit public input.
Grassroots organization is translating into concrete policy wins. Brockovich reports that at least 14 states—including Georgia, Virginia, Maryland, Pennsylvania, New York, Vermont, New Hampshire, Michigan, Indiana, Illinois, Minnesota, Wisconsin, South Dakota, and Oklahoma—are considering full or partial bans or temporary moratoriums on new data center construction. These moratoriums typically range from six months to two years as officials pause to gather more information under public pressure.
On the municipal level, more than 120 cities and towns have enacted some form of pause, moratorium, or outright ban. Notable examples include Monterey Park, California (total ban), and San Marcos, Texas—the first city in Texas to completely prohibit new data centers within city limits. Other victories include the successful halting of the Midtown Project in Milwaukee and, after years of resistance and a key court decision, Blackstone’s QTS abandoning its appeal to build the world’s largest data center in Virginia.
Opposition mounting from local citizens, armed with evidence and testimony, has repeatedly led national and global developers to withdraw contentious projects, as was seen with significant withdrawals in both Milwaukee and Virginia. The pattern is clear: organized, persistent public pressure is increasingly forcing developers’ hands.
Data center resistance is reshaping local politics. Council members and officials who voted in favor of unpopular data center deals are being voted out in places like Festus, Missouri. In some cases, grassroots leaders themselves run for office and use their new positions to put key issues, such as data center bans, to a public vote.
Statewide, bipartisan backlash has even unseated powerful incumbents. The Utah House Speaker lost his seat due to bro ...
Community Resistance Against Data Center Development
Erin Brockovich describes a landscape devoid of regulation or oversight, creating a “free for all” where tech companies and major corporations operate with little accountability to the public or environment.
Brockovich emphasizes that there is no federal oversight, state regulation, or mandatory reporting concerning data center construction, operations, or chemical disposal. Tech companies exploit these regulatory gaps, making decisions that prioritize cost over the well-being of communities and the environment. She shares accounts of whistleblowers who report hazardous waste disposal taking place without proper testing or transparency.
A key example is the PG&E chromium-6 contamination in Hinckley. Brockovich explains how PG&E misled the community by claiming the chromium in the water was a natural element, when in fact it was man-made and highly toxic chromium-6. This deception enabled them to conceal the truth until the statute of limitations was almost up. Critical documents revealed alarmingly high levels (58 ppm) of hazardous chromium-6 being dumped at the facility, but by hiding this information, PG&E was able to downplay the public health risk with later, lower readings. This secrecy allowed them to run out the statutory clock and argue residents could not have gotten sick from supposed “low” exposure, ignoring prior, higher exposures.
Attorney Ed Mazry successfully argued in court that PG&E’s deception reset the statute of limitations, allowing the case to proceed. This argument set a precedent that discovery of corporate cover-ups should restart the statutory countdown, ensuring that deliberate transparency failures do not shield corporations from legal action.
Meanwhile, examples beyond PG&E highlight the lack of executive accountability. The Sackler family of Purdue Pharma faced no criminal charges despite [restricted term]’s role in over 800,000 deaths and billions in illegal profits. Although Purdue was fined $5.5 ...
Corporate Accountability, Regulatory Failures, Federal Legislation Needed
Shawn Ryan and Erin Brockovich discuss the staggering power of major tech players—Google, Meta, Amazon, OpenAI, Grok—estimating their combined wealth surpasses that of most nations, possibly even the United States itself. Nvidia is projected to become the largest company by 2027, symbolizing how AI-driven wealth is consolidating at an unparalleled scale.
This immense corporate growth comes at a price. Erin Brockovich describes a culture where corporate leaders favor cutting costs and maximizing returns for shareholders, even if it means harming the environment and neglecting long-term infrastructure needs. She draws a direct analogy to the Ford Pinto theory: the auto manufacturer weighed the potential expense of fixing a dangerous design flaw versus the anticipated costs of lawsuits and chose the cheaper, more dangerous route. Today’s tech companies apply similar logic, pushing environmental safeguards aside in favor of short-term profits—even as the cost of environmental litigation is ballooning from millions to tens of billions.
Brockovich warns that current business models sacrifice safety and sustainability for immediate financial benefit, running risks "down the line" for later generations to face. She argues upfront investment in safer infrastructure may cost more initially, but yields greater, lasting savings and healthier communities in the long run. As billion-dollar lawsuits over PFAS chemicals and wildfires become commonplace, the argument that it’s "cheaper to face litigation" no longer holds.
Brockovich highlights international models demonstrating that environmentally responsible tech is both possible and already in play. In China, underwater eco-regulated data centers run on the ocean floor, benefiting from strict environmental mandates. Norway is experimenting with Nvidia-powered data centers placed on cold fjords, utilizing wind power for green energy.
New visions for AI infrastructure are on the horizon, too. Brockovich references reports that Nvidia is working on compact, modular data centers—eventually as small as an air conditioning unit on the side of a home—instead of the sprawling, water-hungry facilities that dominate the US. Yet she notes that meaningful adoption of such solutions in the US will require strong, enforceable regulation; without it, companies revert to less responsible, profit-first approaches.
Both Ryan and Brockovich critique the widespread corporate reliance on carbon offsets and greenwashing. Tech companies publicly support carbon offset programs and green energy, but the reality is their data centers often devastate local ecosystems, draining aquifers and disrupting surrounding wildlife—harms that can’t be offset by projects in distant locations.
Brockovich emphasizes that "carbon offsets paying for green projects elsewhere does not restore" damaged local resources. Aquifers, groundwater, forests, and communities harmed by data center expansion aren’t repaired by conservation efforts halfway around the globe. The notion that these offsets constitute real environmental solutions is, she argues, largely a marketing scheme.
AI itself can expose the contradictions within tech’s corporate mindset. Brockovich recounts asking Gemini, an advanced AI model, where it would site new data centers. The AI responded it would choose locations with "zero conflict with human resources," directly critiquing the harm current infra ...
Ai Industry Growth, Corporate Greed, and Environmental Stewardship
Download the Shortform Chrome extension for your browser
