AI-powered software could recover 300GW of hidden capacity for the US power grid, enough to power thousands of AI data centers

AI software claims hidden grid capacity could ease America's growing electricity shortage GridCARE says simulations reveal unused transmission capacity across existing power infrastructure Existing transmission lines may hold far more capacity than previously estimated A new software platform claims it can unlock roughly 300 gigawatts of hidden electrical transmission capacity across the existing United States power grid within three to five years. The technology, developed by GridCARE and led by founder and CEO Amit Narayan, relies on advanced grid modelling rather than costly new infrastructure. Instead of building additional transmission lines or substations, the platform analyzes how the grid actually operates in real time. Unlocking hidden capacity The US power grid has traditionally been planned around conservative assumptions that account for multiple simultaneous equipment failures at once. This approach has left substantial portions of the transmission network underused for most of the calendar year, while electricity demand has resumed growing sharply, and grid upgrades may struggle to keep pace before 2030. Bank of America data indicates the country could face a 100 GW power shortfall within the next four years. Analysts project at least 230 GW of new power demand will emerge between 2026 and 2030 alone. During the same period, utilities are expected to add only 93 GW of new supply capacity. That gap between projected demand and available supply has intensified pressure on operators searching for faster solutions. However, GridCARE claims it could cut years from clean energy interconnection wait times across multiple regions. Running quadrillions of simulations The platform reportedly runs quadrillions of simulations to identify where unused transmission capacity remains hidden from conventional planning tools. By modelling actual grid behaviour rather than worst-case scenarios, utilities gain a more accurate picture of available headroom. This method all