AIPolicyTracker

AI incident ·

xAI Allegedly Operates Unpermitted Methane Turbines in Memphis to Power Supercomputer Colossus to Train Grok

9 news reports Synced from source · record last edited 24 Sep 2026

In brief

An AI system built by xAI and Data center developers and deployed by xAI and Data center operators allegedly harmed South Memphis community, Residents with pre-existing health conditions and 5 others.

Risk domain
Socioeconomic & Environmental Harms Environmental harm
Occurred
Coverage
9 reportsJul 2024 - Mar 2026

What happened

The xAI "Colossus" supercomputer facility in South Memphis allegedly runs 33 methane-powered turbines, despite a permit for only 15. The turbines reportedly power AI training for Grok. The facility's massive energy use and alleged unpermitted turbines reportedly produce extreme nitrogen oxide and formaldehyde emissions, adding toxic smog to a predominantly Black neighborhood already burdened by pollution and health risks.

Editor's notes

Reported timeline of events: (1) 06/05/2024, it is reportedly announced that xAI will begin constructing and operating Colossus, a massive AI supercomputer facility in a predominantly Black neighborhood in South Memphis. (2) Throughout 2024 and into 2025, reports emerge that the facility is running 33 methane-powered gas turbines despite having permits for only 15, reportedly producing large volumes of nitrogen oxides and formaldehyde without public notice or pollution controls.

Laws that address this harm

Policy angle: Classified under Socioeconomic & Environmental Harms (Environmental harm) in the MIT AI Risk Repository taxonomy; 5 recorded instruments address this use case.

Matched from the record's risk domain and country to the instruments recorded here. A reviewer can correct the match in the repository (data/external/incident_overrides.yaml).

News reports (9)

Titles link to the original publisher; report text is not reproduced here.

  1. Elon Musk Gambles Billions in Memphis to Catch Up on AI
    wsj.com · Alexander Saeedy, Kevin Wurm
  2. Inside the Dirty, Dystopian World of AI Data Centers
    theatlantic.com · Matteo Wong, Landon Speers

Who was involved

Alleged deployer
xAI Data center operators
Alleged developer
xAI Data center developers
Alleged harmed party
South Memphis community Residents with pre-existing health conditions Residents living in the vicinity of the Colossus supercomputer Predominantly Black communities General public of Tennessee General public Environmental quality

AI systems implicated

xAI GrokUnpermitted energy infrastructureMethane gas turbinesIndustrial emissions controlEnergy infrastructureData centersColossus

Classification (MIT AI Risk Repository taxonomy)

Causal entity
Human
Intent
Intentional
Timing
Post-deployment
Harm level
—
Sectors
—
Countries
—

Risk entries describing this failure mode

Entries from the MIT AI Risk Repository coded to subdomain 6.6.

  • Excessive landfill

    "Excessive landfill - Excessive disposal of electrical or electronic equipment leading to ecological/biodiversity damage, and disrupting the livelihoods and eroding the rights of local communities."

    A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  • Excessive water consumption

    "Excessive water consumption - Excessive use of water to cool data centres and for other purposes, leading to water restrictions or shortages for local communities or businesses."

    A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  • Biodiversity loss

    "Biodiversity loss - Over-expansion of technology infrastructure, or inadequate alignment of technology with sustainable practices, leading to deforestation, habitat destruction, and fragmentation and loss of biodiversit...

    A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  • Excessive energy consumption

    "Excessive energy consumption - Excessive energy use, leading to energy bottlenecks and shortages for communities, organisations, and businesses."

    A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  • Carbon emissions

    "Carbon emissions - Release of carbon dioxide, nitric oxide and other gases, increasing carbon emissions, exacerbating climate change, and negatively impacting local communities."

    A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  • Electronic waste

    "Electronic waste - Electrical or electronic equipment that is waste, including all components, sub-assemblies and consumables that are part of the equipment at the time the equipment becomes waste"

    A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  • Environmental

    "Environmental - Damage to the environment directly or indirectly caused by a technology system or set of systems."

    A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

  • Natural resource depletion

    "Natural resource depletion - Extraction of minerals, metals, rare earths, and fossil fuels that deplete natural resources and increase carbon emissions."

    A Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation Harms (Abercrombie2024)

Linked by editors or by text similarity in the source dataset.

Other incidents involving xAI

Source record: incident #1144 on the AI Incident Database · all 9 reports