Skip to Content

Meta Breaks Ground on $1B+ AI-Optimized Tulsa Data Center, Expanding US Capacity and Local Workforce Investment

Meta says the Oklahoma project will support AI workloads while pairing expansion with water, grid, and community commitments.

Meta adds another major US AI infrastructure site

Meta announced it is breaking ground on a new AI-optimized data center in Tulsa, Oklahoma, adding another large facility to the company’s global infrastructure footprint. The site is notable not only because it is Meta’s first data center in Oklahoma, but because the company framed it as a direct investment in the compute backbone required for its next wave of AI products and services.

In the company’s announcement, Meta said the Tulsa project represents more than $1 billion in regional investment. At peak construction, the build is expected to support roughly 1,000 onsite jobs, with around 100 permanent operational roles once the facility is fully online. Those numbers matter in today’s AI race, where infrastructure announcements increasingly carry both technology and economic development implications.

Meta also detailed a broader package of community and utility commitments. The company said it plans to invest more than $25 million in local infrastructure improvements, including roads and water systems, and to fund programs that support workforce development for digital infrastructure careers. That includes collaboration with Tulsa Tech and Tulsa Community College, with a target pipeline of graduates in technical trades and data-related disciplines.

On the sustainability side, Meta said the facility is being designed with a closed-loop liquid cooling system intended to minimize water use, and reiterated its broader water-positive goal for 2030. The company also said it will cover data-center utility costs directly and continue matching electricity use with clean energy procurement.

For local governments and enterprise buyers alike, the Tulsa project is another sign that AI infrastructure decisions are now intertwined with policy, utilities, talent, and long-term regional planning. Winning in AI is no longer just about model releases; it is increasingly about who can build reliable physical capacity fast enough and operate it responsibly.

Why it matters

As hyperscalers compete on AI services, data center expansion has become a strategic battleground. Meta’s Tulsa move reinforces that the AI economy is being shaped as much by power, cooling, labor, and permitting execution as by model architecture.

Source: Meta Newsroom

NVIDIA and Google Cloud Expand AI Factory Stack With Rubin Systems and Confidential Blackwell Compute
The partners outlined a deeper infrastructure roadmap for agentic AI and industrial workloads at Cloud Next 2026.