OpenAI details $20 billion Georgia data center, pledges $80 million in local community benefits

OpenAI has published details of Project Camellia, a long-term data center it is designing and developing in Effingham County, Georgia, backed by a 3.2-gigawatt power contract with Georgia Power delivered in phases between 2028 and 2032.

What's new

OpenAI says it is contracting with Georgia Power Company for 3.2 gigawatts of power to support the facility, delivered in phases over roughly four years. Alongside the technical scope, OpenAI laid out specific community commitments: it says electricity rates will not rise for residents because of the project, since OpenAI will pay the full infrastructure and electric-service costs required to serve it, and under Georgia Public Service Commission rules those costs cannot be passed on to existing ratepayers. The company also says the data center will be one of the first designed to proactively reduce its own power draw during periods of high residential demand.

On water use, OpenAI says the project will run a closed-loop system that recirculates rather than continuously withdraws and discharges water, comparing its ongoing water needs to those of an office building with a similar headcount. On community investment, OpenAI committed $80 million in community benefits over the life of the project — directed toward schools, public safety, health care, utilities, workforce training, housing, veterans' services, and small businesses — separate from projected tax revenue. OpenAI says it expects to pay hundreds of millions of dollars in state and local taxes and to become the largest taxpayer in the county, while creating thousands of construction and permanent on-site jobs with a priority on local contractors. It is also making up to $71 million in Codex credits available to eligible Georgia college, community college, and technical school students, with $100 in credits per eligible student through their ChatGPT accounts.

Context

Project Camellia is part of OpenAI's broader buildout of dedicated data center capacity to meet training and inference demand, following a string of similar large-scale infrastructure commitments this year. The detailed, upfront community-benefits framing — addressing electricity rates, water use, taxes, and jobs directly — reflects growing local pushback nationally against data center projects over their strain on power grids and water supplies, which OpenAI is addressing preemptively rather than after opposition emerges.

Why it matters

At 3.2 gigawatts, Project Camellia is a large addition to the physical infrastructure underpinning OpenAI's model training and serving capacity, and the accompanying $80 million community-benefits package plus education-credit program is a notable template for how a frontier AI lab is choosing to negotiate local acceptance for a data center at this scale. As power and water availability increasingly gate how fast AI labs can expand compute, the specific mechanics here — a closed-loop water system, proactive demand curtailment, and a public commitment that ratepayers won't subsidize the build — are likely to be referenced by other developers and by communities negotiating their own data center deals.