OpenAI has published the National Community Compact and a Compute Campus Policy Blueprint, two documents that define how the company will build and operate future AI compute campuses in the U.S. The compact applies to campuses that OpenAI owns and operates and commits to creating a local compact for each site, engaging residents early enough that input shapes project plans, providing a local point of contact, and reporting publicly each year on performance. The blueprint is aimed at federal, state, and local policymakers and proposes rules for the broader industry, arguing that voluntary company commitments are a starting point but that new policy is required to build trust between communities, elected officials, and companies.

The compact includes three common commitments. First, OpenAI says it will pay the full cost of the electricity it uses and the infrastructure needed to serve it, and will provide financial assurances before major spending so other customers are protected if a project is delayed, reduced, or canceled. Second, it will develop a water plan for each campus, prioritize closed-loop or low-water cooling, publish site-level water and electricity use, publish construction plans and independent noise studies before operations and annually, and avoid NDAs with public officials except for narrow protections covering security information, trade secrets, or proprietary/competitively sensitive information. Third, it will publish expected jobs, work with unions and training providers to prepare residents, report annually on hiring and apprenticeships, and make sustained investments in priorities that residents help select.

The post gives two examples of community-shaped investments. In Effingham County, Georgia, residents helped shape a $4.8 million grant to provide free school meals for all 14,000 county students. In Pike County, Ohio, families identified childcare as a priority, leading to a joint $6 million commitment from OpenAI and SB Energy.

The policy blueprint asks state utility commissions to require large electricity customers to cover the costs they add to the grid and to protect other customers if a project falls through. It also asks utilities to make it easier for campuses to connect if they can safely cut demand when power is scarce. The post cites the planned Effingham County campus as an example: an agreement with Georgia Power allows the utility to reduce delivered power by up to 1,000 megawatts during high demand. It suggests Congress can help by speeding up transmission upgrades and restoring and expanding tax incentives for clean electricity.

For accountability, the blueprint calls for public reporting on water, energy, noise, and hiring, narrow limits on required NDAs, and public terms for state or local tax incentives, with measurable commitments and a way to recover benefits if a developer falls short. To keep work local, it proposes sharing purchasing plans early so local suppliers can compete for contracts, and using hiring forecasts so employers, building trades unions, and community colleges can train residents for jobs ranging from electrical construction work to long-term maintenance of power and cooling systems.

The announcement frames these steps as a continuation of work begun in January with the Stargate community commitments to pay their own way on energy and develop plans shaped by local priorities, followed by commitments in Georgia and Ohio covering energy costs, water use, local investment, and accountability. The core message is that building compute infrastructure in the U.S. can lower AI delivery costs and expand access, and that communities hosting campuses should have influence, cost protection, and a share of economic gains. The compact applies only to campuses OpenAI owns and operates; the blueprint is a proposal for policymakers, not a binding standard.