Meta’s massive data center bet is a direct challenge to OpenAI and Alphabet

Meta may not currently lead the race for AI superintelligence, but it’s drawing heavily on its cash reserves to pursue the technology. Founder Mark Zuckerberg announced Monday that the company will spend hundreds of billions of dollars to build several enormous AI data centers.

The first of these centers—expected to be online next year—is called “Prometheus.” Zuckerberg says it will be capable of generating one gigawatt of energy. But Meta isn’t stopping there.

“We’re also building Hyperion, which will be able to scale up to 5GW over several years. We’re building multiple more titan clusters as well. Just one of these covers a significant part of the footprint of Manhattan,” he wrote on Facebook.

Five gigawatts is roughly equal to the total energy consumption of Miami and equivalent to the output of five nuclear reactors—enough to power about 3.5 million homes for a year.

There’s a certain irony in naming the center Prometheus. In Greek mythology, Prometheus defied the gods by stealing fire from Olympus. Meta, meanwhile, has been aggressively poaching top AI talent from rivals, offering multi-million-dollar salaries that far exceed what competitors pay. Zuckerberg is reportedly overseeing those recruitment efforts himself. With data centers of this scale, he could further strengthen the company’s hiring appeal.

Meta has also acquired a 49% stake in Scale AI, bringing its CEO, Alexandr Wang, on board. Nat Friedman, former CEO of GitHub, has also joined Meta. Daniel Gross—the CEO and co-founder of Safe Superintelligence, which he launched with OpenAI co-founder Ilya Sutskever last June—is now part of the company as well.

With this wave of infrastructure announcements, Zuckerberg is signaling Meta’s intent to outpace OpenAI and other artificial intelligence initiatives. This comes after the company’s Llama 4 model underperformed following several high-profile staff departures. In response, Meta reorganized its AI division and renamed it Superintelligence Labs.

And Meta certainly has the funds to support this push. In April, it raised its expected capital expenditure to as high as $72 billion. On Monday, Meta’s stock rose just under 1% to $724 per share. The company currently has a market capitalization of $1.82 trillion—vastly higher than OpenAI’s $300 billion valuation. (Meta didn’t immediately respond to Fast Company‘s request for comment.)

But the competition isn’t sitting by either. OpenAI is building a five-gigawatt data center of its own, called Stargate. Announced in January, the company said it would invest $500 billion over four years to build the facility, with support from Oracle, SoftBank, and MGX.

OpenAI committed to deploying $100 billion immediately. Texas has been designated as the flagship location for the data centers, with the first site expected to begin operations later this year in Abilene.

“This project will not only support the re-industrialization of the United States but also provide a strategic capability to protect the national security of America and its allies,” OpenAI wrote in a blog post.

Alphabet, meanwhile, is spending $3.3 billion on two new data centers in South Carolina.

One issue Meta did not address: the environmental impact of these massive data centers. A scientific paper published last year on Cornell University’s preprint server arXiv (titled “The Unpaid Toll: Quantifying the Public Health Impact of AI”) estimated that pollution from AI data centers could cause up to 1,300 premature deaths annually by 2030. It also estimated that public health costs related to their air pollution are already at $20 billion per year.

By 2030, the researchers forecast, the public health burden from AI data centers will be twice that of the U.S. steel industry, and could rival emissions from all the cars, buses, and trucks in California.

Meanwhile, the Trump administration has dismantled dozens of climate programs in its first 100-plus days. The Environmental Protection Agency (EPA) is also considering overturning previous findings that classify greenhouse gas pollution as harmful—potentially weakening its ability to regulate carbon emissions.