A House Democrat Wants Musk to Answer for xAI’s Polluting Data Centres
July 30, 2026 – 10:47 am
Image by: Congressman Frank Pallone
Frank Pallone, the top Democrat on the House Energy and Commerce Committee, is demanding that Elon Musk account for the pollution from xAI’s data centres, escalating a fight over the gas turbines that power the company’s supercomputers.
Demands and Complaints:
Pallone sent a letter on July 29th seeking:
- Emissions data
- Permitting records
- Access for committee staff to visit the data centre sites
The facilities in question are xAI’s Colossus clusters, located in Tennessee and Mississippi, which house some of the largest AI training systems globally. These centres rely on natural gas turbines installed to meet their substantial power demands, a rapid expansion driven by the need to train models like Grok, xAI’s alternative to ChatGPT.
The Focus: Gas Turbines
The heart of the complaint lies in the 59 gas turbines installed at Colossus 2 without securing federal clean-air permits, as required for such significant emissions. Pallone accused the company of "disregard for the health and well-being of local communities" after a meeting where xAI’s representatives provided insufficient information.
Location and Environmental Justice:
The location near predominantly Black neighborhoods around Memphis and across the state line intensifies the controversy, transforming an air-quality dispute into an environmental-justice issue. The NAACP has already announced plans to sue over the operation, and Pallone’s letter adds to growing legal and political pressure on xAI and Musk.
Defenses and Arguments:
xAI and the Justice Department have argued that the turbines are vital for national security, as its Grok models are used in military targeting. However, this argument raises a permitting question from a defense perspective. The entity is currently folded into Musk’s wider empire and reported to be owned by SpaceX, leading to Musk being directly targeted with these demands.
Broader Tensions:
The dispute highlights a larger tension between AI’s compute demands and clean power availability. As companies compete to train larger models, they resort to gas turbines and diesel generators, betting on future advancements in renewable energy to catch up.