Musk’s xAI to Build World’s Largest AI Site with New Power Permit

Date:

In a landmark decision, Mississippi state regulators have approved a permit for Elon Musk’s xAI to build a 41-turbine power plant in Southaven. This facility will provide the energy needed to run the company’s expanding AI supercomputing infrastructure, including the Colossus 2 site. The decision is a key part of Musk’s goal to create the world’s most powerful single-site AI training installation.
The 41 methane gas turbines will generate approximately 1.2 gigawatts of power, enabling xAI to train its Grok models at an unprecedented scale. This “behind-the-meter” power generation is a direct response to the massive energy requirements of modern AI hardware. By building its own power source, xAI can compress development timelines that would otherwise take years.
The project has faced fierce opposition from local residents and the NAACP, who argue that the state is “rubber-stamping” a major polluter. Community members have expressed concerns about the noise and air quality issues associated with the turbines. The NAACP has filed a notice of intent to sue, alleging that xAI is violating the Clean Air Act with its “copy and paste” turbine strategy.
Health advocates warn that the emissions from the turbines, including nitrogen oxide and formaldehyde, will disproportionately affect local families. Both DeSoto and Shelby counties have already received failing grades for air quality from the American Lung Association. The Southern Environmental Law Center (SELC) has called for a more thorough review of the project’s long-term environmental impacts.
While the legal battle unfolds, xAI is moving forward with the $20 billion “Macrohardrr” datacenter, its third in the region. This massive investment is expected to transform Mississippi into a global hub for AI innovation. However, the tension between this economic boom and the environmental health of the local community continues to grow.

Related articles

Mark Zuckerberg Closes the Book on Metaverse After Racking Up Nearly $80 Billion in Losses

Meta is drawing a line under one of the most costly experiments in corporate history. Horizon Worlds, the...

Instagram Drops Privacy Shield: What the End of DM Encryption Really Means

Meta's decision to strip end-to-end encryption from Instagram direct messages before May 8, 2026 is being described by...

Google’s Health AI Feature Offering Stranger-Sourced Medical Tips Is No Longer Active

Google has discontinued a search feature that used AI to surface health recommendations gathered from ordinary internet users....

Microsoft Doubles Down on Support for Anthropic as Pentagon Stands Firm on Unprecedented AI Blacklist

As the Pentagon stands firm on its unprecedented decision to blacklist Anthropic, Microsoft has doubled down on its...