On Nov. 2, 2026, OpenAI launched the Grid Protectors Program, a national framework for future compute campuses to cut electricity use when regional grids are under strain. The first concrete numbers sit in Georgia, where OpenAI's planned Effingham County campus—Project Camellia—is paired with a Georgia Power agreement that includes up to 1,000 megawatts of flexible demand response.
Key facts
| Item | Detail |
|---|---|
| Program | Grid Protectors |
| Announced | Nov. 2, 2026 |
| Pilot | Effingham County, Georgia (Project Camellia) |
| Flexible demand | Up to 1,000 MW |
| Campus need | About 3,200 MW per Georgia Power |
What OpenAI announced
OpenAI describes Grid Protectors as “a national effort to support affordable, reliable electricity for families and businesses in the communities where we build.” Working with utilities, regulators and policymakers, it will “develop arrangements for our future campuses to reduce electricity demand when the grid is under strain.” Those arrangements will cover how much demand can be reduced, how quickly and for how long. OpenAI says they will differ by location, and that it “will share additional commitments as they are finalized.”
It cites a projection that U.S. data center capacity will exceed 100 gigawatts by 2030 and wants flexibility baked into campuses it has yet to build.
How the flexibility works
Depending on local needs and equipment, OpenAI may also work with grid operators to supply power from on-site generation or storage, and use batteries and power controls to smooth changes in demand.
OpenAI argues that grid infrastructure is a fixed cost whether demand is low or high. Using more power when the grid has room and less when it is strained can spread those costs and limit peak-driven upgrades—a rationale separate from Georgia Power's contractual caps.
The Georgia pilot and Georgia Power numbers
For the planned Effingham County campus (Project Camellia), OpenAI committed to pay the full infrastructure and electric-service costs. The campus is still in development, but OpenAI intends Effingham to be the Grid Protectors pilot.
Georgia Power expects the facility to need approximately 3,200 megawatts. Under a 25-year agreement, the utility can reduce energy delivered during certain times to support grid stability. OpenAI's side includes up to 1,000 megawatts of flexible demand response—roughly the average demand of 700,000 households, by OpenAI's comparison.
Georgia Power calls the deal “among the largest single-facility demand response commitments in the country” and says it can invest in fewer new generation resources to serve growth. OpenAI will meet long-term energy contract requirements and financial assurances to protect Georgia Power customers.
OpenAI cites Georgia Power projections of approximately $950 million in annual customer savings beginning in 2029 from its portfolio of large-load agreements and expected growth, including Project Camellia, plus at least $15 a month for a typical residential customer using 1,000 kilowatt-hours. Those savings figures are Georgia Power projections cited by OpenAI.
Who pays and the cost argument
OpenAI states that “Families and small businesses should not be asked to subsidize a compute campus's electricity needs.” It says large computing facilities should pay their costs—including infrastructure to serve campuses—while helping utilities manage the systems they depend on.
Georgia Power's release aligns: OpenAI pays full infrastructure and service costs and backs long-term contracts and assurances so other customers are protected.
What it does not do
OpenAI warns that “Flexibility cannot replace investment in new energy supply and transmission.” Beyond Georgia, it has not finalized site-by-site reduction commitments; additional deals may come as arrangements are completed.
What it means
Analysis (ours). Multi-gigawatt AI sites face a political test: prove neighbors will not subsidize peak load. Grid Protectors bundles demand response, optional on-site supply and full cost recovery into a template Effingham is meant to pilot. The gap between about 3,200 MW of expected need and up to 1,000 MW of curtailment shows how utilities may swap flexible load for fewer peaking plants. That parallels the silicon push in Microsoft Unveils Maia 300, Plans Multi-Gigawatt Rollout—chips and power contracts are both bets that AI scale can stay grid-compatible.
What to watch
Final Grid Protectors terms for campuses outside Georgia, operational proof of up to 1,000 MW curtailment as Project Camellia advances, and whether other regulators demand similar packages. Track whether Georgia Power's projected 2029 savings hold as more large-load deals stack up—and whether critics accept flexibility without parallel supply and transmission builds.