Data centers expanding to support artificial intelligence workloads are facing significant delays in securing connections to the power grid and acquiring new gas turbines, sometimes waiting as long as seven years for 100-megawatt hookups. To accelerate deployment timelines, some operators are turning to an unconventional solution: repurposed jet engines from retired aircraft, modified to run on natural gas. These jet engine generators can be installed rapidly, allowing a data center to become operational within roughly two years.

Notable companies using this strategy include Elon Musk’s xAI, which has deployed eight of these converted turbines in Mississippi to boost power at its Memphis Colossus 2 facility. OpenAI followed suit with gas turbines at its Stargate site in Texas, while Crusoe, a data center startup, ordered 13 units earlier this year for its AI campuses. These engines, typically GE models formerly used on Boeing jets, are rebuilt by ProEnergy and provide a swift if imperfect power source amid infrastructure bottlenecks.

However, experts warn that these jet turbines come with trade-offs such as lower efficiency and increased pollution compared to conventional gas turbines designed for power plants. Using more engines to generate similar output raises environmental concerns, suggesting the approach is more of a short-term fix than a sustainable strategy. Industry views indicate that efficiency and emissions will likely become pivotal considerations within the next couple of years as data centers explore long-term power solutions.

Policy developments under the Trump administration encouraging data centers to self-generate electricity add another layer of complexity to the mix. Despite these pushes and the urgency of current AI-driven growth, the transition from makeshift jet-engine power to greener, more reliable infrastructure is expected to remain a critical challenge for data center builders as they navigate extended waiting periods for traditional energy resources.