Musk's 2 Colossal AI Data Centers Burn Natural Gas: 3 Stocks to Watch

The AI arms race has a new front: energy independence. Musk's two Tennessee data centers won't wait for grid upgrades; they're drawing power from a privately built natural gas plant in Mississippi. For AI builders and investors, this signals that fossil fuel infrastructure is now a critical enabler of rapid model training and deployment. The stocks of pipeline giants like Kinder Morgan may ride the AI wave as much as chipmakers.

Elon Musk is tackling the AI industry's most pressing infrastructure bottleneck—reliable, scalable power—by going off-grid with natural gas. Two colossal AI data centers under construction in Tennessee, dubbed Colossus I and II, will be powered by a dedicated natural gas plant in Mississippi, bypassing the congested and politically charged electricity grid. This move, revealed in late July 2026, has immediate implications: it reinforces natural gas as the pragmatic bridge fuel for the AI era, draws legal pushback from residents and regulators, yet reportedly enjoys U.S. government backing. For investors, it highlights midstream pipeline operators Enterprise Products Partners, Enbridge, and Kinder Morgan as potential beneficiaries of a new demand driver.

“

For investors, it highlights midstream pipeline operators Enterprise Products Partners, Enbridge, and Kinder Morgan as potential beneficiaries of a new demand driver.

The choice of fossil fuel for AI infrastructure is no anomaly. Tesla's recent earnings detailed heavy spending on AI, and SpaceX's upcoming first public earnings report is expected to underscore similar commitments. AI data centers require enormous, uninterrupted power—often 30-100 MW per facility—and grid interconnection queues can stretch three to five years. Musk's solution, purchasing natural gas turbines and building a plant on-site, exemplifies a broader industry paradigm: speed-to-power is becoming as critical as compute density. Natural gas offers relatively quick construction, high reliability, and a lower capital outlay than nuclear or large-scale renewables paired with storage. It also sidesteps the electricity price inflation that grid-connected data centers impose on residential and commercial consumers—a growing pain point for regulators.

Yet the environmental and social costs are immediate. The Mississippi power plant faces lawsuits from nearby communities, and environmental advocates warn that locking in new natural gas capacity for AI could derail net-zero pledges. However, the U.S. government's reported support suggests a high-level prioritization of AI dominance over climate concerns in the near term. This tension encapsulates the modern energy dilemma: the very technology touted to optimize everything from grids to supply chains is itself a voracious emitter.