🎥 Watch the full interview here

Read the full report here:

Read the full report here:

Watch the AI Data Center Ecology webinar:

In the spring of 2026, residents of San Marcos, Texas, came out for the second time to oppose a proposed $1.5 billion data center project at 904 Francis Harris Lane.

And they won.

After hours of public testimony, the San Marcos City Council voted 5–2 against the land-use change needed for the project to move forward.

We believe this is one of the most important data center case studies in the United States.

It reflects a community-wide effort to protect an area connected to crucial Texas Hill Country water systems, including the Edwards Aquifer, San Marcos Springs, the Blanco River, and the Guadalupe River.

And there is something we can learn from it.

By following the San Marcos story from 2025 through 2026, we were able to recognize some of the important steps developers have to take before a hyperscale data center becomes operational—and where communities may have an opportunity to get involved.

So, if you’re wondering whether a data center could be coming to your area, here are some things to watch for.

Subscribe now

1. Site Selection

A hyperscale data center needs access to power—and a lot of it.

That’s why one of the first things to look for is proximity to major electrical infrastructure: high-voltage transmission lines, substations, and large power plants.

In the San Marcos case, the proposed site was located next to the Hays Energy Power Plant and high-voltage transmission infrastructure.

Purchasing land near power infrastructure doesn’t guarantee that a data center will actually receive the electricity it needs. But it can make a location significantly more attractive to developers.

So if you see hundreds of acres being purchased near major transmission infrastructure, pay attention.

It might be nothing. But it’s worth asking questions.

2. Going Industrial

The next clue may appear in a place most people never think to look: zoning documents.

Once land is purchased, a developer may need to change its zoning or land-use designation before a data center can be built.

Agricultural, conservation, residential, or undeveloped land may suddenly be proposed for commercial or industrial use.

This is where things get interesting.

Rezoning and land-use changes can require public hearings, depending on the jurisdiction and the type of approval being requested. These meetings can give residents an opportunity to learn about a project, submit comments, and speak on the official record.

And sometimes, this is where a community has the most leverage.

In San Marcos, the proposed project required changes to the city’s Preferred Scenario Map and zoning before it could move forward. The fight over those changes ultimately became the obstacle the project couldn’t overcome.

Know Your Jurisdiction

Before you can follow a project, you need to know who actually has authority over it.

Depending on the location, a proposed data center may fall:

• Within city limits

• Within an Extraterritorial Jurisdiction (ETJ), where applicable

• Outside city limits, on county or unincorporated land

Why does this matter?

Because it determines which government bodies you need to watch.

Inside a city, that may mean the city’s planning and zoning commission and City Council. Outside city limits, county authorities may play a larger role. ETJ powers vary by state and locality.

Following the San Marcos story taught us something important: Don’t wait until construction equipment arrives.

By then, some of the most important decisions may already have been made.

3. Follow the Money

Another place to look is tax incentives.

There are two major types of tax benefits commonly associated with data center development:

1. State-level sales and use tax exemptions

2. Local property tax abatements

These agreements can reveal a lot about a project before most people have even heard about it.

Take CyrusOne’s DFW10 project in Bosque County, Texas. The county approved a ten-year agreement waiving 30% of certain county property taxes. The agreement also placed responsibility for construction-related road damage on CyrusOne and encouraged the use of local vendors—but did not impose a strict local-hiring requirement.

And that’s a relatively modest example.

Louisiana provides qualifying data centers with sales and use tax exemptions for eligible equipment and development expenses.

So if you want to know whether a large project could be coming to your community, don’t only watch zoning agendas.

Watch the tax deals.

Sometimes the money trail tells the story before the construction site does.

4. Follow the Power

Being located next to a transmission line doesn’t automatically mean a data center can plug into it.

Large facilities generally have to secure electrical service and, when connecting directly to the transmission system, go through an interconnection process involving utilities and grid operators.

And right now, those systems are under enormous pressure.

In December 2025, PJM—the regional transmission organization operating the largest U.S. power grid—held its capacity auction for the 2027–2028 delivery year.

The result?

PJM’s came up 6.6 GW short of its 20% installed reserve margin target.

Data center developers are keeping their options open by applying for multiple grid interconnections across mutiple electric grids, even if they’re not intending to build.

This has been causing a massive inflation in forecasting and wholesale electricity prices.

This is why power agreements, interconnection requests, new substations, transmission upgrades, and proposed power plants are all worth watching. They can reveal what is coming.

When the Grid Isn’t Enough

Developers are also finding other ways to power enormous facilities.

Some data centers are being built next to existing power plants. Others sign long-term Power Purchase Agreements (PPAs) with electricity producers.

And some are going even further—building their own power generation.

That can mean large fleets of natural gas turbines operating almost like a private power plant beside the data center.

Which brings us to another stage communities should watch very closely.

5. Environmental Permits

A data center isn’t only a building full of servers.

Large facilities can require backup generators, cooling infrastructure, water infrastructure, substations, and—increasingly—on-site power generation.

Those systems can trigger environmental permitting requirements, including air and water permits.

And those permits create another paper trail.

Look for applications filed with state environmental agencies, county health departments, air-quality agencies, water authorities, and other regulators.

Let’s learn from xAI’s Colossus 1 and Colossus 2 to see how things can unfold—and what communities should watch for. Read the full case study in Chapter 11 of our Report.

Whether you support or oppose a particular project, the lesson is the same:

Environmental permits can tell you what a data center actually plans to operate—not just what the project looks like on a rendering.

Learn to Recognize the Signs

A data center doesn’t appear overnight.

Before the walls go up, there can be land purchases, zoning applications, annexation requests, tax agreements, interconnection studies, power contracts, environmental permits, public hearings, and government votes.

Individually, these things may look boring. Together, they tell a story.

And the earlier communities recognize that story, the more time they have to ask questions.

• Who is the developer?

• How much electricity will the facility need?

• Where will the water come from?

• Will the land need to be rezoned?

• Are tax breaks being offered?

• Will new transmission infrastructure be required?

• Will there be gas turbines or hundreds of backup generators?

And, perhaps most importantly:

When is the public allowed to speak?

San Marcos showed that community involvement can matter. Hundreds of people showed up, local organizations mobilized, thousands signed a petition, and residents stayed at City Hall into the early hours of the morning before the final vote.

The project was stopped, for now.

That doesn’t mean every data center can be stopped—or that every project follows exactly the same approval process. But it does mean communities don’t have to wait until construction begins to start paying attention.

Spotting a project early is the first step, having the knowing where else to look is the second, and acting is the third.

Share

Want to learn how a hyperscale data center moves from an idea to a fully permitted facility—and where communities can get involved along the way? Read Chapter 11 of our Report.

Download the AI Date CenterReport

Read the rest of our Data Center stories here.

The detailed post we published in February about digital identity and identification systems is still more than relevant. It covers digital ID rollouts across North America, Europe, and Asia, identifies the key beneficiaries, outlines the main risks, and discusses possible solutions.

Even though data centers are dominating the conversation right now, digital identity shouldn’t be forgotten. It adds another layer of pressure to our everyday lives, which are already surrounded by anti-privacy and surveillance technologies, devices, and legislation.

Share

The Idea Isn’t New. The Speed Is.

The idea of digital identity itself isn’t new. We can trace its roots back to the 1970s, with more concrete forms beginning to take shape in the 2000s.

What is new is the speed.

We’re witnessing rapid changes and a steady rollout of new legislation, with different approaches and timelines depending on the region. The direction, however, remains largely the same: encouraging people to adopt digital IDs and normalizing the idea of having a digital identity.

In The Article, We Talk About:

• What is digital identity?

• What is a digital ID?

• The history of digital identification and the future timeline

• Global efforts to introduce digital IDs

• The European digital identity timeline

• Digital ID in the U.S.: Apple, Google, and TSA

• Digital identity in Asia

• How digital identity ties together digital ID, digital money, travel, and health systems

• Who benefits from the digital ID rollout?

• Cybersecurity threats and concerns about freedom

• And most importantly: solutions to the digital identity problem

Get the Full Picture

This Is Just the Beginning

Digital identity is moving fast, and understanding where it’s heading matters. We put everything you need to know into a free eBook—download it and see the bigger picture.

🎥 Watch the full interview here

Read the full report here:

Read the full report here:

Watch the AI Data Center Ecology webinar:

Read the full report here:

Watch the AI Data Center Ecology webinar:

Our water is in danger. As data centers multiply, they add a new, thirsty tenant to already stressed aquifers. When groundwater drops, wells run dry, streams shrink, and the ecosystems and communities that depend on them suffer.

This isn’t an exaggeration. History—and the data—confirm that it’s already happening.

The United States has experienced serious groundwater declines for decades as a result of intensive development and water use, particularly from agriculture, industry, and growing communities.

A 2024 USGS study of global groundwater trends across 1,693 aquifer systems worldwide concluded that the groundwater level is declining rapidly—up to 1.5 feet every year in many regions.

Those declines translate into fewer gallons for private and agricultural wells, rivers, wetlands, and the wildlife they sustain.

Adding Data Centers to an Existing Problem

Now a rapidly growing industry is demanding huge water volumes for cooling from an already unstable water system. And the affects of that won’t just stop at the data center property lines.

Hyperscale data centers can consume millions of gallons of water for cooling, depending on their cooling technology, climate, size, and operating conditions. As these facilities become larger and more concentrated, their water demand can place additional pressure on local water supplies.

Water levels can fall. Wells can become less productive or run dry. Groundwater-dependent ecosystems can be affected. And communities can be left competing for the same limited resource.

This is especially concerning in regions that are already experiencing water stress.

Subscribe now

What About Water Quality?

Quantity isn’t the only concern.

Groundwater can become contaminated through both natural processes and human activities like industrial discharges, urban activity, agriculture, groundwater pumping, and waste disposal.

Pumping water can change the way it naturally moves underground, adding another delicate variable.

This creates a dangerous combination: less water and greater vulnerability to contamination.

In some circumstances, polluted water at or near the land surface can infiltrate into groundwater recharge areas. If groundwater flow is altered by pumping, contaminants can be transported toward wells and drinking-water supplies where it was never an issue previously.

The Problem Is Bigger Than the Data Center

The question is simple: How much additional pressure can our water systems absorb?
And who decides where to draw the line?

When a hyperscale data center arrives in a community, its water requirements become part of a much larger system of wells, aquifers, rivers, municipalities, farms, homes, and ecosystems.

Water doesn’t belong to an industry, it’s a shared resource necessary for our survival. Once an aquifer is depleted or contaminated, restoring it can take years, decades, or, in some cases, not happen at all. That’s a big deal.

That’s why data center expansion cannot be treated as simply another construction project.

Before any new facility is approved, communities need to know:

1. Where the water will come from

2. How much will be used

3. How the aquifer will be replenished

4. What happens to surrounding users when demand spikes

Because once the water is gone, no amount of corporate sustainability language can bring it back.

Download the AI Data Center Report

Read more about our water system analysis in Chapters 8 and 9 of our Report, and use our AI Data Center Map to estimate water displacement near you by clicking on an empty space on the map.

Share

Follow the series. Read the it here.

🎥 Watch the full interview here

Read the full report here:

Read the full report here:

Watch the AI Data Center Ecology webinar:

Share

GrapheneOS is the most secure mobile operating system available today. Built on an open-source foundation, it offers complete transparency; eliminating the risk of hidden backdoors and ensuring true trustworthiness. Its advanced security architecture includes a hardened kernel, minimized attack surface, and app permission controls that put the user firmly in charge.

The key difference between these two phones lies in the security update timeline. The Pixel 8 Pro is supported with security updates through October 2030, while the newer Pixel 9 Pro extends that support to August 2031.

Even in the worst-case scenario, such as theft or confiscation, your data remains protected. Whether you’re using a Pixel 9 Pro, Pixel 8 Pro, or any Above Phone, your information is shielded by the uncompromising strength of this digital fortress.

Considering how similar these two models are, we decided to publish a detailed comparison to help our customers understand the differences. Here’s what we covered:

• Display advancements

• Camera innovations

• Battery life comparison

• Design differences and weather resistance

• Performance comparison

• Side-by-side comparison table.

  Read More

🎥 Watch the full interview here

Read the full report here:

Watch the AI Data Center Ecology webinar:

By reading big tech's environmental reports, analyzing them with common sense and some basic math, we found that major data center developers cannot be trusted to share accurate and complete information.

While we spent the last decade being warned about climate change, big tech is now presenting us with carefully crafted sustainability narratives—filled with uplifting visuals of new jobs, clean energy, social programs, and environmental initiatives surrounding data centers.

Taken together, we found numerous inconsistencies in both what these companies report and how they report it. In some cases, important information appears to be omitted altogether.

• There is no clarity on which data centers are metered and which rely on estimates.

• They hide behind efficiency metrics while avoiding disclosure of raw resource consumption.

• Their sustainability reports highlight water conservation projects without explaining how aquifers are naturally replenished through precipitation and water infiltration from rivers, lakes, and ponds.

Reading the appendix of Google’s 2025 Environmental Report, we find that water usage across Google’s facilities is sometimes estimated. In some cases, it isn’t measured at all.

Lawrence Berkeley National Laboratory estimates that fewer than one-third of U.S. data center operators measure their water consumption. A separate 2021 study by Uptime found that 50% of surveyed data centers tracked water use in some way, while fewer than 10% measured water consumption across their entire fleet.

There’s no excuse for avoiding actual measurements—which, if publicly disclosed, might reveal the true impact data centers have on local water resources.

These companies have an undeniable obligation to measure and disclose their use of water—our planet’s most precious resource.

Subscribe now

Sustainability Report Breakdown

We’ve reviewed the sustainability reports published in 2025, covering resource usage during 2024, so you don’t have to. Here’s a summary.

Google — Google publishes its total electricity consumption (32 TWh), total water withdrawal (11 billion gallons), and power and water usage by site. However, it’s unclear how many facilities report measured values versus estimated ones.

Meta — Meta’s report showcases carbon emissions, renewable energy, and biodiversity initiatives. You need to dig into the Environmental Data Index to find its total electricity use (18 TWh) and total water consumption (5.625 billion liters). The report also breaks down power and water usage by site, but it’s unclear how many facilities report measured values versus estimates.

Microsoft — Microsoft publishes global totals: 30.207 TWh of electricity consumption and 10.377 billion liters of water use. According to the company, most of its water comes from third-party suppliers, primarily local municipalities.

Amazon — Amazon publishes a sustainability report but provides neither total electricity consumption nor water usage. Instead, it focuses primarily on efficiency metrics rather than actual facility-level consumption.

On the other hand, we were unable to find recent environmental or sustainability reporting from X, Anthropic, or OpenAI at the time of writing.

After going private in 2022, X was no longer obligated to publish environmental disclosures. Historically, Anthropic and OpenAI rented computing capacity from cloud providers. As of 2026, however, both are developing data center capacity of their own.

One Golf Course, Two Golf Courses…

If you think all of this is unacceptable, listen to this.

Google uses the term “golf course equivalents” to describe the water impact of each data center.

A reader might reasonably ask:

“How much water could a golf course actually use?”

Reading the fine print, we find that Google is referring to large golf courses in the arid southwestern United States—courses that require more water than almost anywhere else in the country.

An average golf course in that region can consume around 150 million gallons of fresh water per year.

Does this make Google’s data centers look better?

Not really.

It makes golf courses look worse.

Who Is Holding Them Accountable?

The organizations responsible for managing local water resources often have limited authority to enforce meaningful restrictions.

There is a massive accountability gap when it comes to large-scale water users, and it’s unlikely we can rely solely on local water authorities to prevent abuse.

It would be naive to expect someone else to protect our water. We need to speak up, demand transparency and accurate reporting, and oppose data centers that are draining our water resources.

Download the Report

Read more about big tech’s sustainability reporting in Chapter 7 of our Report, and learn how our water system works in Chapter 8.

If you come across additional data on data center electricity or water consumption, please share it with us using our contact form.

Share

Want the full picture? Explore the rest of the series here.

🎥 Watch the full interview here

Read the full report here:

Read the full report here :

Watch the AI Data Center Ecology webinar:

Ethical AI is artificial intelligence designed and used in a way that respects human values, privacy, and autonomy.

Share

To help people understand the bigger picture, we wrote an article explaining the fundamentals of AI technology and why it should be viewed as a tool—not an autonomous thinking organism.

In the Article, You’ll Learn:

• What AI actually is

• All about machine learning and deep learning

• How AI is trained

• What a large language model (LLM) is

• The difference between AI models and AI agents

• What data centers are and what they need to operate

• The differences between cloud AI and ethical AI

• Resources to help you easily set up your own AI

We also created a free webinar on this topic, which you can watch at any time after signing up for a free account. And if it sparks your curiosity, be sure to check out our next webinar, where we explore AI data centers in greater depth.

Read More

🎥 Watch the full interview here

Read the full report here:

Read the full report here:

Watch the AI Data Center Ecology webinar: