OpenAI launches Project Camellia AI data center in Effingham County, Georgia

OpenAI announced Project Camellia, a new AI data-center build in Effingham County, Georgia. Beyond the compute itself, the announcement emphasizes community-facing commitments: responsible energy sourcing, local economic investment, job creation, and providing the community with access to Codex, OpenAI's coding agent. The framing mirrors the playbook large AI builders now use to win local support and permitting for power-hungry facilities.

The mechanism is capacity: as OpenAI's model tiers proliferate (GPT-5.6 Sol/Terra/Luna) and it touts 10M ChatGPT Work and Codex users, it needs enormous incremental compute, and geographic diversification of data centers helps with power availability and grid load. Georgia offers land, energy, and incentives that make it attractive for hyperscale AI infrastructure.

Competitively, Project Camellia is another data point in the week's dominant compute-buildout theme, alongside AMD's 2GW commitment to Anthropic, Microsoft's European infrastructure funding for Mistral, and NVIDIA's national AI factories. It also complements OpenAI's parallel push into national science via a Department of Energy partnership. Skeptics point to the environmental and grid strain of AI data centers and note that community-benefit language is now standard PR accompaniment to facilities that consume large amounts of power and water. Watch for disclosed capacity, power sourcing specifics, and whether the Codex-for-community pledge translates into real local uptake.