Industry

Amazon's New Gas Plant Could Be One of the Largest Emitters in the US

Amazon is building a natural-gas power plant in Texas to power its data center, with a permit allowing up to 33 million tons of CO2 emissions – more than the largest coal plant in the U.S. The project, GW Ranch, could become one of the biggest greenhouse gas emitters in the country, highlighting Big Tech's shift to fossil fuels for AI energy demands. This move challenges Amazon's Climate Pledge goal of carbon neutrality by 2040.

Neura News

Neura News

Neura Market Editorial

August 8, 20266 min read
Amazon's New Gas Plant Could Be One of the Largest Emitters in the US

Amazon is investing in a new natural-gas power plant in Pecos County, Texas, built to serve its own data center. The project, called GW Ranch, could become one of the largest single producers of greenhouse gases in the United States.

The plant is designed with 35 natural-gas turbines. Those turbines would deliver 7.65 gigawatts of electricity, primarily to Amazon's new data center on the same site. Initially, the plant would not be connected to the state's power grid.

A Permit That Exceeds Coal

Texas regulators granted GW Ranch a permit allowing emissions of up to 33 million tons of CO2. That single permit allows more carbon dioxide than the largest coal plant in the country. Cleanview, an organization that tracks data centers and associated power projects, provided the data behind the report, which was first published by the New York Times.

Plants rarely emit as much greenhouse gas as their permits allow. Even so, the scale of the allowance is striking. The permit's lax pollution restrictions are notable, according to the source material, and the plant's emissions could undermine Amazon's climate goals.

Amazon confirmed it purchased the site and plans to buy power from GW Ranch. The company did not dispute the permit figures. The project sits in West Texas, far from major population centers, but its emissions would be felt globally.

Big Tech's Fossil Fuel Turn

Amazon is not alone in this direction. Meta and Google have started building their own power plants. Across the industry, companies are turning to gas, coal, and other non-renewable energy sources to meet the skyrocketing energy demands of AI data centers.

The shift is significant. Data centers need power around the clock, and renewables alone have not kept pace with demand. The Trump administration has sought to lift restrictions on polluting power plants, making it easier for companies to pursue fossil-fuel projects.

The source material describes this as a new heyday for gas. The economics favor speed and reliability, even at the cost of higher emissions. For Amazon, the move is a sharp departure from its public environmental commitments.

The Climate Pledge Under Pressure

Jeff Bezos, Amazon's founder, made the Climate Pledge commitment. The pledge, co-founded by Amazon and Bezos, set a target to make the company carbon neutral by 2040. That target now looks harder to reach.

Amazon's emissions have climbed for several years in a row due to AI demands. The company's data centers use enormous amounts of electricity and water. A related article noted that Amazon's data centers used 2.5 billion gallons of water last year alone.

The GW Ranch plant could make the 2040 goal even more distant. Even if the plant only emits a portion of its permitted gases, it would be a significant setback for the Climate Pledge. The permit alone signals a willingness to accept high emissions in exchange for AI capacity.

Amazon's Response and What the Plant Means

Amazon spokeswoman Margaret Callahan commented on the situation to the New York Times. She said, "The world looks different now than when we co-founded the climate pledge," while adding that Amazon's commitment to the pledge hasn't changed.

The #1 Newsletter in AI

Stay ahead of the AI curve

The most important updates, news, and content — delivered weekly.

No spam. Unsubscribe anytime.

The comment acknowledges the tension. The pledge was made in a different energy landscape. Now, AI's appetite for power is reshaping corporate decisions. Callahan did not address the specific permit or the plant's potential emissions in the quoted remarks. The company has not said whether it will offset the plant's output or pursue other mitigation measures.

The GW Ranch project is a concrete example of how AI is changing energy markets. A single data center can now justify a dedicated power plant with more capacity than many utilities. The 7.65 gigawatts from the turbines would be enough to power a mid-sized city.

The permit allows more CO2 than the largest coal plant in the country, a striking comparison. Coal has long been the target of environmental regulation, yet this gas plant's allowance exceeds it. Gas burns cleaner than coal per unit of energy, but the sheer scale here is what matters.

The plant would initially operate off-grid. That means it would not be subject to the same market pressures or grid reliability rules. It would run for one customer: Amazon's data center.

The article was published on Aug 8, 2026, as part of a series on AI data centers. The author is Terrence O'Brien, weekend editor at The Verge. The series has examined how the AI boom is driving energy consumption and infrastructure decisions.

A Broader Pattern

Amazon's move fits a wider trend. Tech companies are no longer just buying power from utilities. They are becoming power producers themselves. Meta and Google have both started building their own plants, and the pattern is likely to continue.

The Trump administration's efforts to ease restrictions on polluting power plants have made these projects easier to permit. Federal policy has shifted toward energy availability over emissions reduction. That shift is visible in projects like GW Ranch.

The plant's location in Pecos County is also telling. West Texas is known for wind and solar, but this project runs on gas. The choice of fuel reflects the need for reliable, dispatchable power that renewables cannot always provide.

For Amazon, the challenge is balancing growth with its stated climate goals. The company has invested heavily in renewable energy, but AI demands are outpacing those efforts. The GW Ranch plant is a direct response to that pressure.

The 2040 target remains on the books. Whether it survives contact with projects like this one is another question. The permit alone suggests a willingness to accept emissions that would have been unthinkable a few years ago.

The full impact of the plant will depend on how much it actually runs. Permits set ceilings, not actual output. But even partial operation would add millions of tons of CO2 to the atmosphere each year.

Amazon has not said how it plans to reconcile the plant with its pledge. The spokeswoman's comment suggests the company sees the pledge as adaptable. The world has changed, and so have the company's energy needs.

The GW Ranch project is now a test case. It shows how far tech companies will go to secure power for AI. It also shows how quickly climate commitments can bend under that pressure.

Related on Neura Market

More from Neura News

Industry

Google Cuts Pixel 11 Pro AI Trial to Six Months, Adds Three Costly Catches

Google has reduced the free Google AI Pro trial bundled with the Pixel 11 Pro from 12 months to six months, cutting the perk's value by $119.94. The change applies across the Pixel 11 Pro lineup and introduces three costly catches, including losing the trial if upgrading to AI Ultra, auto-renewal before the next flagship launch, and termination of existing promos when redeeming new ones. The Pixel 10 Pro still offers the full 12-month trial, making it a viable alternative for shoppers.

Aug 16·4 min read
Research

LittleLearner Models Trained Only on K-5 Curriculum Show Skills Are Elicited, Not Acquired

Researchers released LittleLearner, a family of language models trained from scratch on a strictly filtered K-5 elementary school curriculum, to answer whether capabilities beyond training data can be elicited or acquired through scaling, post-training, and in-context learning. The answer is largely no: scaling, post-training, and in-context learning amplify what the curriculum taught, but none meaningfully improve out-of-scope performance. The pretraining filter sets the effective capability ceiling, providing a controlled sandbox for studying knowledge acquisition and RL.

Aug 16·5 min read
Industry

The Hidden Gold Rush: Scammers Exploit Demand for Claude Watermark Removal Apps

Anthropic's August 2026 watermarking of Claude text has sparked a surge in demand for removal apps, attracting scammers who peddle fraudulent tools. AI scientist Lance Eliot warns these apps often contain malware or fail to work, as statistical watermarks are nearly impossible to remove without heavy editing. With billions of users at risk, the problem is expected to worsen as more AI makers adopt watermarking.

Aug 16·12 min read