Tech & AI
Data Centers Are Burning Fossil Fuels. It’s Destroying Big Tech’s Climate Goals
From Ashburn, Virginia, to South Dublin, Ireland, the data centers that journalist Paris Marx visited all looked remarkably similar: large industrial facilities with loud air-conditioning units running to keep racks upon racks of servers cool. But it was the security measures around the data centers that struck Marx more than anything.
Tall fences. Security cameras everywhere. Locks on every door that required one of the few full-time employees to unlock with a badge. And in one memorable instance, standing on an Irish public curb outside a cluster of Microsoft-owned centers, there was an eerie feeling of being watched by men on lawn mowers, trimming grass that clearly didn’t need to be cut.
“It was just really staggering to me that it was at that level that it was so hard to move through these spaces,” Marx told me.
The intense security should give you an idea of just how serious Big Tech is about its data centers. These facilities are the key to the artificial intelligence programs at OpenAI, Meta, Google, Amazon and elsewhere, creating vast amounts of compute and sucking up equally staggering amounts of electricity and water.
There is no stronger metaphor for the manifestation of Big Tech’s power than the increasing sprawl of data centers. Amazon, Microsoft and Google alone control 59% of hyperscale data center capacity. In his new book available now, Hyperscale: The Ambition and Excess of Big Tech’s Data Empires, Marx, host of the Tech Won’t Save Us podcast, highlights how global data center protests are one of many ways that communities are pushing back against Big Tech’s power over us.
Hyperscale demands
Data centers existed long before generative AI; we called them server farms, and they’re what keep us online. But in this AI era, we have hyperscale data centers.
Before, a large data center might require around 50 megawatts of power to keep everything running. Now, hyperscale data centers can require triple that amount of power. Newer ones, like Meta’s Louisiana data center, may require 4,500 megawatts (4.5 gigawatts) — staggering demand that officials say is four times more than all of the electricity used in New Orleans on its hottest day in 2025. It’s more than the two gigawatts the Hoover Dam produces.
Hyperscale data centers are also housing more advanced tech. Chips like GPUs are key to creating compute, and newer chips built for AI data centers can run hotter. That makes it even harder to cool facilities, so in addition to air conditioning, facilities are using water systems to avoid servers overheating. It’s an additional environmental toll. A hyperscale center can suck up 1 million to 5 million gallons of water every day, equivalent to a town of more than 10,000 residents, according to a 2025 estimate.
So how do you keep these facilities running? The process to get approved and connected to a town’s electrical grid can take years. So tech companies have started building their own power plants right next to data centers for the sole purpose of keeping them on — perfect for burning massive amounts of fossil fuels, the main contributor to the climate crisis.
Hyperscale impacts
In the 2010s, Amazon, Microsoft, Google and many others pledged some version of becoming carbon neutral (or negative) in the next few decades — meaning they would offset the emissions they created with at least as much emission-reducing actions.
Sustainability reports from that time show that, overall, tech companies were measurably decreasing their environmental footprint. “And then ChatGPT comes along, and all of a sudden, that’s out the window,” Marx told me.
Tech leaders have been candid about allowing generative AI to override their climate goals. Google reported a 50% increase in its greenhouse gas emissions in 2024 to support its generative AI. Microsoft’s pledge to be carbon negative by 2030 “was before the explosion in artificial intelligence,” Brad Smith, the company’s president, told Bloomberg in 2023, saying the carbon goal was “five times as far away as it was in 2020” because of AI.
Modern data centers and their power plants burn fossil fuels for power, requiring gas turbines, huge tanks of diesel fuel kept on-site in case of a power outage or good old-fashioned coal. This has devastating environmental and health consequences, both immediately for local communities and in the long term for the entire world.
Communities that neighbor data centers have reported brown or unsafe drinking water, unending noise, and in one lawsuit against Elon Musk’s xAI centers in Memphis, extreme health risks; Memphis residents “are feeling it in their lungs because of the fossil fuels that are being burned,” Marx said.
Tennessee state representative Justin Pearson speaks at a rally in opposition to Elon Musk’s data center in Memphis on April 25, 2025.
Brandon Dill for The Washington Post via Getty Images
And as decades of climate research prove, burning additional fossil fuels is only going to exacerbate the effects of climate change. That’ll lead to even more extreme weather, health risks and natural disasters. As Marx put it in his book, “The world will suffer through the worsening natural disasters fueled by the quest for computation by some of the most powerful people on the planet.”
With data centers and their fossil fuels, we’re fundamentally reshaping the planet for AI. So anti-data center movements are asking: Is it worth it?
“How much computation do we really need here?” Marx asked in our conversation. “How many resources are we willing to dedicate to this vision of the future? And is this technology really delivering in such a way that expending all those resources is justified?”
The answer — for many data center neighbors in Memphis, Oregon, Chile, Ireland and thousands of other communities across the globe — is an unflinching no.
You must be logged in to post a comment Login