OpenAI has launched a programme it calls Grid Protectors, under which its future compute campuses are meant to draw less electricity when power grids are under strain. The company announced it on 2 November 2026 and named Georgia as the first place it will apply.
The Georgia project is a planned campus in Effingham County. OpenAI has an agreement there covering up to 1,000 MW of flexible demand response, which it says equals the average demand of about 700,000 households. Georgia Power said in July that the facility is expected to need about 3,200 MW under a 25-year agreement that lets the utility reduce the energy it delivers at certain periods.
OpenAI says it will work with utilities, regulators and policymakers on how much demand is cut, how fast and for how long. It may also supply on-site generation or storage, with batteries to smooth demand swings. It has committed to pay the full infrastructure and electric-service costs for the Effingham project.
The company's argument is that fixed grid costs are shared across more sales when demand is steady, and that trimming peaks limits the upgrades a grid needs, which keeps costs down for other customers. Georgia Power says demand response lets it build fewer new generation resources. It projects about $950 million a year in customer savings from 2029 and a benefit of at least $15 a month for a typical home using 1,000 kWh, though those figures come from its whole large-load portfolio, not the OpenAI campus alone.
OpenAI adds that flexibility cannot replace new energy supply and transmission investment. It will share more commitments as they are finalised, and Effingham remains a pilot still in development.
Sources · 1