As artificial intelligence reshapes the boundaries of human computation, it is also reshaping the boundaries of energy conscience. The tech industry's long-cultivated image of green data centers is straining under the weight of AI's insatiable power demands, pushing giants like Meta, Amazon, and Microsoft toward natural gas — and toward Exxon Mobil, which is now positioning carbon capture technology as the moral bridge between fossil fuel reality and climate aspiration. In advanced negotiations with hyperscalers, Exxon is offering to capture 90 percent of emissions from natural gas plants powe
Exxon Pursues Carbon Capture Deal to Power AI Data Centers with Natural Gas
probably the only realistic game in town to accomplish that
Mark
Why would a tech company that's built its brand on sustainability turn to natural gas in the first place?
Mimi
Because renewable energy can't keep up. AI training runs 24/7, and the sun doesn't always shine. Nuclear is the dream, but plants take a decade to build. Natural gas is there, it's proven, and it works right now.
Mark
So Exxon is basically offering them a way to have it both ways—fossil fuels without the carbon guilt?
Mimi
That's the pitch. Capture 90 percent of the emissions and you can tell your shareholders you're still committed to decarbonization. Whether that actually works at scale is a different question.
Mark
What's the catch for Exxon?
Mimi
They get to stay relevant. Instead of being phased out, they become the infrastructure provider for the AI age. It's a long-term play—they're betting the tech companies will take the deal because they have no better option.
Mark
And if carbon capture doesn't work as advertised?
Mimi
Then everyone loses. The tech companies look hypocritical, Exxon's credibility takes a hit, and the climate problem gets worse. But right now, nobody's willing to wait for the perfect solution.
The Pulse
- AI's computational hunger has outpaced the renewable energy infrastructure built to feed it, forcing tech companies to quietly abandon green commitments in favor of reliable natural gas.
- The gap between public climate pledges and operational reality is widening fast — Meta has already signed a natural gas deal in Louisiana, signaling that the industry's green facade is cracking under pressure.
- Exxon CEO Darren Woods declared his company 'probably the only realistic game in town' for decarbonizing natural gas at scale, a bold claim that reflects both corporate confidence and the absence of ready alternatives.
- Carbon capture technology — expensive, energy-intensive, and never deployed at the scale now being proposed — is being asked to carry the weight of an entire industry's climate credibility.
- If Exxon closes these deals, natural gas with carbon capture becomes legitimized as the de facto energy strategy for the AI era, potentially locking in fossil fuel infrastructure for decades while nuclear and renewables race to catch up.
As artificial intelligence reshapes the boundaries of human computation, it is also reshaping the boundaries of energy conscience. The tech industry's long-cultivated image of green data centers is straining under the weight of AI's insatiable power demands, pushing giants like Meta, Amazon, and Microsoft toward natural gas — and toward Exxon Mobil, which is now positioning carbon capture technology as the moral bridge between fossil fuel reality and climate aspiration. In advanced negotiations with hyperscalers, Exxon is offering to capture 90 percent of emissions from natural gas plants powering AI infrastructure, a proposition that is less a triumph of clean energy than a reckoning with how difficult the clean energy transition truly is.
The artificial intelligence boom is colliding with the limits of the tech industry's green energy strategy. Training and running AI models demands constant, uninterrupted electricity — something solar and wind, by their intermittent nature, cannot reliably provide. Nuclear power is promising but years away from meaningful new capacity. So major hyperscalers like Meta, Amazon, and Microsoft are turning to natural gas, and Exxon Mobil is stepping forward with a proposition: burn the gas, but capture the carbon.
On a recent earnings call, Exxon CEO Darren Woods described advanced negotiations with power providers and tech companies to deploy carbon capture at natural gas plants supplying AI data centers. The target is capturing 90 percent of CO2 emissions — an ambitious figure that would, if achieved, dramatically reduce the climate footprint of facilities that are otherwise difficult to decarbonize quickly. Meta's recent deal with Louisiana utility Entergy to power a data center campus with natural gas illustrates just how real this shift already is.
Exxon's argument rests on what it already owns: decades of carbon capture expertise, existing infrastructure for transporting and storing CO2, and the ability to act now rather than in some future decade. Woods was candid about the competitive landscape, suggesting Exxon faces little near-term rivalry in this specific niche.
The deeper stakes, however, are harder to resolve. Carbon capture works in principle but has never been proven at the scale now being contemplated, and it is both costly and energy-intensive to operate. Whether captured carbon remains permanently sequestered is a question science has not yet fully answered at industrial scale. For Exxon, success here means new revenue and extended relevance in an era that might otherwise leave fossil fuels behind. For tech companies, it means data centers can grow without waiting for a cleaner future to arrive. For the climate, it is a wager — on technology, on execution, and on whether the least bad option available is good enough.
The artificial intelligence boom is hungry for power, and the tech industry's renewable energy strategy is hitting a wall. Major companies like Meta, Amazon, and Microsoft have long marketed their data centers as green operations, but the relentless computational demands of training and running AI models are forcing them to look elsewhere—and increasingly, that means natural gas. Now Exxon Mobil is positioning itself as the solution to a problem the tech sector created for itself: how to burn fossil fuels without the guilt.
During an earnings call on Friday, CEO Darren Woods outlined the company's pitch to hyperscalers—the industry term for the handful of giants building massive data center campuses. Exxon is in advanced negotiations with power providers and technology companies to deploy carbon capture technology at natural gas plants that feed electricity to AI infrastructure. The goal is ambitious: capture 90 percent of the carbon dioxide emissions from those facilities. "I'm hopeful that many of these hyperscalers are sincere when they talk about the desire to have low emission facilities," Woods said, "because certainly in the near to medium term we're probably the only realistic game in town to accomplish that."
The statement reveals the bind that tech companies find themselves in. Renewable energy—solar, wind, hydroelectric—has been the public-facing answer to data center power needs for years. But renewables are intermittent, and AI training requires constant, reliable electricity. Nuclear power is emerging as a second option, with companies making major new investments in atomic energy. Yet nuclear plants take years to build and face regulatory hurdles. Natural gas, by contrast, is available now, proven, and already embedded in the nation's power infrastructure. Meta exemplifies this shift: the company recently signed a deal with the Louisiana utility Entergy to power a data center campus with natural gas rather than renewables.
Exxon's pitch is that it can make this compromise palatable. The company has spent decades developing carbon capture technology—extracting CO2 from industrial processes and either storing it underground or using it for other purposes. Woods emphasized that Exxon possesses the existing infrastructure, the technical expertise in capturing and transporting carbon dioxide, and the storage capacity to make the model work. The company is not starting from scratch; it is repurposing capabilities it already owns.
What makes this moment significant is not the technology itself but the market signal it sends. If Exxon succeeds in inking deals with major tech companies, it would effectively legitimize natural gas as a long-term energy source for AI infrastructure—provided the carbon is captured. For Exxon, it opens a new revenue stream and extends the relevance of fossil fuels into the AI era. For tech companies, it offers a way to scale data centers without waiting for nuclear plants to come online or betting entirely on renewable capacity that may not materialize fast enough. For the climate, it is a more complicated calculation: carbon capture technology works, but it is expensive, energy-intensive, and has never been deployed at scale. Whether 90 percent capture is achievable in practice, and whether the captured carbon stays sequestered, remains an open question.
Woods's confidence that Exxon is "probably the only realistic game in town" suggests the company believes it has time before competitors emerge. But the statement also hints at the urgency driving these conversations. The hyperscalers need power now. They are not waiting for a perfect solution. They are looking for the least bad option available, and if Exxon can deliver reliable electricity with a carbon capture backstop, that may be enough to close the deal.
Notable Quotes
In the near to medium term we're probably the only realistic game in town to accomplish that— Darren Woods, Exxon CEO
Want the full story?
Read the original at CNBC ↗
Technology