Meta's Energy Thirst: Why the AI Giant Is Betting Big on Natural Gas

The construction of Meta's Hyperion data center reveals a paradox: the company is expanding its AI infrastructure with natural gas plants, challenging its sustainability goals in the face of colossal energy demand.

Meta's Energy Thirst: Why the AI Giant Is Betting Big on Natural Gas
AI in Business
1 de April de 2026
3

The race for supremacy in Artificial Intelligence has reached a level where electrical consumption is no longer measured in megawatts, but on the scale of entire states. The latest example of this voracious appetite is the Hyperion project, a new Meta data center that, when fully operational, will require an electrical load equivalent to the total demand of South Dakota. To support this massive operation, the tech giant announced the funding of seven new natural gas plants, adding to three others already planned. Together, these 10 plants in Louisiana will generate 7.5 gigawatts, exceeding the energy capacity of the entire aforementioned US state.

The paradox of corporate sustainability

For years, Meta has positioned itself as a pillar of environmental responsibility in the tech sector. The company frequently highlights its sustainability reports and robust investments in renewable sources, even facilitating the operation of a nuclear power plant for two decades. However, the decision to bet on natural gas puts these promises under a rigorous microscope. While gas is often defended as a 'transition fuel' — a temporary solution while clean energies and storage technologies mature — this justification has proven increasingly fragile in the face of technical and climate reality.

Technical challenges and the methane problem

From a technical standpoint, the choice of natural gas raises immediate concerns about emissions. Calculations based on data from the U.S. Department of Energy indicate that the turbines in Louisiana will release 12.4 million metric tons of CO2 annually. This volume represents 50% more than Meta's entire carbon footprint in 2024. The problem is compounded by methane, the main component of natural gas, which has a global warming potential 84 times greater than carbon dioxide. Even leakage rates of just 0.2% in the supply chain can make the climate impact of gas greater than that of coal, while the average leakage rate in the U.S. is around 3%.

Impact on climate goals

Meta's silence on methane in its most recent sustainability reports is a critical point of concern. The company, which has historically led purchases of solar and wind energy, now faces a dilemma: how to balance the extreme need for power for advanced language models with its commitment to carbon neutrality. The scale of the Hyperion project is so vast that the use of fossil fuels could permanently compromise the company's decarbonization trajectory, forcing it to rely excessively on carbon removal credits to offset the operation, a strategy that many experts consider insufficient given the magnitude of the actual emissions.

Competitive context and the energy market

While the price of renewable energy and storage batteries has plummeted over the last decade, the cost of gas turbines has been on an upward trajectory. Meta's choice is therefore atypical and complex, especially because other tech companies have diversified their investments into baseload power sources, such as small modular nuclear reactors. Meta is, in practice, choosing a path of high emission intensity at a time when the AI industry should be seeking innovations that decouple computational growth from accelerated environmental degradation.

The future of AI infrastructure

What we can expect from here on is increasing pressure on Meta to provide an honest and transparent accounting of its methane emissions. The Hyperion project is not just data processing infrastructure; it is a stress test for the ESG (Environmental, Social, and Governance) policies of the entire tech industry. The future will require giants like Meta not only to buy renewable energy but also to take direct responsibility for the integrity of the energy supply chains that underpin their data empires; otherwise, the cost of artificial intelligence may become unaffordable for the planet's climate.

Advertisement
Share
Comments (0)

Sign in to leave your comment

Sign In

Don't have an account? Create account