Elon Musk’s clean energy dream just hit a snag. His AI company, xAI, is running data centers on natural gas turbines — billions of dollars’ worth of them. That’s right. While Tesla sells solar panels, Musk’s newest venture is doubling down on fossil fuels.

The contradiction is glaring. Tesla was founded to kill hydrocarbons and push a solar-electric economy. Musk’s first Master Plan declared this mission clearly. Yet xAI’s power comes from dozens of unregulated natural gas turbines, with plans to buy $2.8 billion more. It’s a fossil fuel commitment, not a temporary fix.

Meanwhile, SpaceX is betting on a different future: space-based solar power. The company claims solar arrays in orbit can generate five times more energy than Earth-based panels, thanks to constant sunlight. Musk envisions launching gigawatts of servers into orbit, powered by these arrays, to run AI operations.

This vision comes with massive challenges. Launch costs remain high, and protecting sensitive hardware from radiation and space hazards isn’t cheap or simple. Plus, it’s unclear if AI training can spread effectively across satellites without high-speed, low-latency links.

For now, xAI’s gas-powered data centers serve as stopgaps. Musk seems confident that orbital data centers will eventually replace terrestrial ones — and with them, the gas turbines. But that’s a big bet on a technology still years away from viability.

The scale Musk imagines is staggering. SpaceX’s IPO filing talks about terawatt-scale annual AI compute growth. To put that in perspective, all the world’s data centers today consume about 40 gigawatts. Musk’s projection implies AI will need a terawatt of power every year — roughly a quarter of all global continuous energy use.

This is Musk’s classic first-principles approach. He identifies a bottleneck — energy supply — and designs a radical workaround. But the assumption that AI compute demand will grow exponentially without limit is unproven. The energy crunch might materialize — or it might not.

Meanwhile, terrestrial solar power’s potential remains underexploited. Shipping solar panels by truck uses far less energy than launching them into orbit. Solar manufacturing on Earth is scaling up, costs have dropped 90% in the past decade, and deployment is faster than any space-based alternative.

The irony is stark. Tesla’s solar and battery business raked in $2.8 billion in just Q1 2026. Yet xAI, a Musk company, relies on natural gas turbines while spending nearly $700 million on Tesla Megapacks for peak load management. Solar panels from Tesla Energy are barely part of the equation.

This split highlights a strategic pivot. Musk hasn’t abandoned solar power outright, but terrestrial solar now seems secondary to a grander, riskier space-based power vision. The clean energy narrative driving Tesla’s early years is competing with a reality that embraces fossil fuels — at least for the immediate future.

Behind the scenes, the compute arms race is brutal. xAI lost $6.36 billion in 2025 alone. Powering massive AI models demands raw energy and uptime, which natural gas turbines provide more reliably for now. Local opposition and grid constraints make terrestrial expansion tricky.

Legal and environmental issues complicate matters too. xAI faces lawsuits over emissions from its gas turbines in Memphis. Still, Musk is doubling down, ordering billions more in turbines to keep pace with AI’s surging demand. This commitment signals fossil fuels won’t disappear from Musk’s operations anytime soon.

The gamble is clear: Musk is chasing a future where orbital solar arrays power AI servers free from Earthly limits. Until then, fossil fuels keep the lights on. Whether the space solar dream can scale, survive costs, and deliver on its promise remains an open question — one with billions on the line.