Dozens of planned natural-gas power plants in the U.S. may emit greenhouse gases on a scale comparable to entire nations [1], [3].

This surge in emissions threatens global climate goals as the digital infrastructure required for artificial intelligence and cloud computing grows. The reliance on fossil fuels to power the AI revolution creates a conflict between technological advancement and environmental sustainability.

Reports indicate that the potential annual CO2 emissions from these new gas-powered data centers could exceed 129 million tons [2]. This volume of pollution is described as being as much greenhouse gas annually as the total emissions from Australia or France [1].

These projects were reported in 2024, with various permits being reviewed and approved throughout 2024 and 2025 [2], [3]. The expansion is driven by the rapidly growing electricity demand of AI, cloud computing, and other digital services [1], [4].

To meet this energy need, operators are turning to natural gas to ensure a steady power supply. However, the scale of the planned infrastructure, comprising dozens of plants [1], has drawn significant attention from climate activists. These groups said that the carbon footprint of the digital economy is becoming a major source of climate-change-linked emissions [1].

While data center operators seek to maintain the reliability of their services, the environmental cost is becoming harder to ignore. The transition to green energy for these facilities remains a critical challenge as the demand for processing power continues to spike.

“Potential annual CO2 emissions from new gas-powered data centers could exceed 129 million tons.”

The scale of these emissions highlights a growing tension between the rapid deployment of generative AI and international carbon-reduction commitments. As data centers move from being passive energy consumers to drivers of new power plant construction, the tech industry's 'net-zero' claims face increased scrutiny from regulators and environmental advocates.