Inside Meta’s $200 billion bet on a remote AI data center

In rural northeast Louisiana, far from Silicon Valley, Meta is making one of the biggest infrastructure bets of the AI era. The company is backing a gigantic new data center campus in Richland Parish, a poor farming community that suddenly finds itself at the center of a $200 billion gamble on artificial intelligence.

What Meta is building in Louisiana

Meta is planning what could become its largest and most ambitious AI data center complex to date, in Richland Parish, Louisiana. The project started as a $10 billion investment but is now expected to reach around $200 billion over time as the campus expands and fills with AI hardware.

The facility is designed to support Meta’s long-term AI ambitions, including training and running massive models that power products like recommendation systems, generative AI features, and future AR/VR experiences. To do that, Meta wants to bring together around a million GPUs in one place—an unprecedented level of compute in a single campus.

A data center the size of a city

The scale of the project is hard to grasp. The planned data center is expected to draw about 7.5 gigawatts of power, supplied by 10 newly built natural gas plants. For comparison, that’s roughly equivalent to the electricity consumption of New York City on a winter day.

In terms of land, the site has been compared to the size of Manhattan when superimposed on a map. Instead of skyscrapers and apartments, it will be filled with warehouse-like buildings, cooling systems, substations, and the racks of servers and GPUs that run Meta’s AI workloads.

Why Richland Parish?

Richland Parish is not the kind of place most people associate with cutting-edge technology. It’s a rural area named for its fertile land, historically tied to cotton plantations and now focused on crops like soybeans and corn. A quarter of its residents live below the poverty line, and the region has struggled with a long cycle of economic hardship.

So why build one of the world’s most powerful AI campuses here? Part of the answer is politics and permitting. As data centers face growing pushback in more densely populated or environmentally sensitive areas, big tech companies are looking for regions where local leaders are eager to clear red tape and offer incentives. The American South has become a prime target.

Entergy Louisiana, the local utility, saw an opening. At the start of 2024, it pitched Meta on the idea that Louisiana could deliver not just huge amounts of power, but also fast execution—getting approvals, land, and infrastructure in place at speed. That combination of energy, land, and political will helped put Richland Parish on Meta’s shortlist.

The tax breaks that sealed the deal

Behind the scenes, Meta made it clear that it would only move forward if the state agreed to major tax breaks, especially on the expensive AI hardware that would fill the data center—GPUs, networking gear, and other equipment.

Because lawmakers couldn’t easily pass a brand-new bill at that point in the legislative calendar, state leaders quietly folded Meta’s requested exemptions into an unrelated rural broadband bill. The result was a package of incentives that dramatically reduced Meta’s tax burden.

In return, Meta promised to create 300 jobs initially, at about 150% of the state’s average annual salary—roughly $82,000 a year—with a plan to reach 500 jobs later to unlock the full incentives. That would translate into about an 80% reduction in property taxes on the broader facility. Crucially, there’s no requirement that those jobs go to local residents, and if Meta misses the hiring targets, the deal doesn’t disappear; it just trims the incentives.

Power, politics, and AI

While the data center deal was taking shape, Louisiana’s political influence in Washington was rising. Representative Mike Johnson unexpectedly became Speaker of the House, and Representative Steve Scalise, also from Louisiana, held the role of House Majority Leader. Together, they gave the state an outsized voice in Congress.

At the same time, Senator Marsha Blackburn of Tennessee was pushing the Kids’ Online Safety Act (KOSA), a bipartisan bill aimed at holding social media platforms—including Meta—more accountable for harms to minors. The bill sailed through the Senate in the summer of 2024, and many expected it to move quickly in the House.

Instead, Speaker Johnson raised concerns about free speech and did not bring the bill to the floor, slowing its progress. Critics questioned whether Meta’s massive investment in Louisiana might have influenced the political calculus, though Johnson publicly denied any connection between the data center project and his stance on KOSA. As of now, the bill still hasn’t passed.

This intersection of AI infrastructure, corporate incentives, and national policy echoes broader trends across the AI industry. As companies race to secure compute and power, their investments can reshape local politics and national regulation—similar to the way other major AI bets, like Elon Musk’s $250 billion push with xAI, are influencing where talent and capital flow.

How Meta is financing the $200 billion buildout

Mark Zuckerberg didn’t just want size and speed; he also wanted financial flexibility. Rather than loading Meta’s own balance sheet with massive amounts of debt, the company turned to Wall Street.

Meta sparked a months-long bidding war among big financial players to finance the construction. The winning structure is a joint venture where Meta owns only about 20% of the campus. The rest is controlled by private capital firm Blue Owl, backed by roughly $27 billion in debt from asset managers including PIMCO.

This setup lets Meta keep the data center off its balance sheet while still retaining control over day-to-day operations. It’s a sign of how AI infrastructure is becoming its own asset class, attracting the same kind of institutional money that chases airports, pipelines, and large real estate portfolios. Similar dynamics are playing out across the AI supply chain, from cloud compute to chips and memory, as seen in moves like Micron’s expansion of US memory production to feed AI demand.

Promises of jobs and a “generational” transformation

For many people in Richland Parish, Meta’s arrival feels like a once-in-a-lifetime opportunity. Local leaders talk about the project as “generational” and “transformational,” hoping it will bring new jobs, higher wages, and a broader economic revival.

In the short term, the region is already experiencing a classic boom-town effect. Construction workers, engineers, and contractors are pouring in. Land values are rising. Some residents imagine the area growing to rival cities like New Orleans in economic importance.

But there’s also anxiety. A modern AI data center is incredibly capital-intensive but not very labor-intensive. Even at full scale, 300 to 500 permanent jobs is a small number compared to the size of the investment. And with no requirement to hire locally, there’s a real risk that many of the highest-paying roles will go to imported talent rather than long-time residents.

From cotton and corn to data

The project is also forcing a deeper question: what does it mean for a farming region to pivot from crops to compute?

Fields that once grew corn, soybeans, or cotton are being reimagined as land for server halls, cooling towers, and power infrastructure. Instead of harvests, the output is data and AI models. For some, that’s an exciting evolution—a way to plug a rural community into the digital economy. For others, it feels like a loss of identity and a gamble on a technology wave that might not deliver lasting local benefits.

Residents are asking what happens if the AI boom slows, or if Meta shifts its strategy in a decade. Will the community be left with stranded infrastructure, empty buildings, and a landscape permanently altered for a promise that never fully materialized?

What’s really at stake

Meta’s Richland Parish project is more than just another data center. It’s a snapshot of how the AI race is reshaping geography, politics, and local economies.

On one side, there’s the potential for high-paying jobs, new tax revenue (even after incentives), and a sense that a long-neglected region is finally part of a global technology story. On the other, there are trade-offs: heavy reliance on natural gas, massive power consumption, opaque financial structures, and political dynamics that may or may not align with the community’s long-term interests.

As AI companies continue to scale up, more towns and regions will face similar choices. Richland Parish is an early test case for what happens when the world’s biggest tech platforms bring trillion-dollar ambitions to places that have been economically overlooked for generations—and whether those ambitions truly lift local communities, or simply use them as staging grounds for the next wave of AI infrastructure.