Meta CEO has revealed the company plans massive AI data centre investments

Meta CEO Mark Zuckerberg announced on Monday that the company will channel hundreds of billions of dollars into constructing multiple huge AI data centres as part of its race toward superintelligence.

This bold move underscores Meta’s determination to lead the next wave of machine-driven innovation, following an intense contest for top AI talent. The social media heavyweight has joined peers in making headline-grabbing deals and handing out multi‑million‑dollar pay packages, all in an effort to speed up the development of systems that could outperform people on a broad array of tasks.

Zuckerberg targets 2026 for first multi‑gigawatt facilities

Zuckerberg revealed on his Threads feed that the inaugural data centre, dubbed Prometheus, is slated to become operational in 2026. A second site, named Hyperion, will eventually scale up to around five gigawatts of power over the coming years. He added that “We are building multiple more titan clusters as well. Just one of these covers a significant part of the footprint of Manhattan,” highlighting the sheer magnitude of the project.

Pointing to industry watcher SemiAnalysis, Zuckerberg noted that Meta is on track to be the first lab to bring an AI supercluster exceeding a gigawatt online. If realised as planned, this feat would mark a major milestone in large‑scale machine learning infrastructure, potentially setting a new benchmark for competitors.

To calm investor jitters about such immense spending, Zuckerberg emphasised Meta’s robust advertising revenue, enough to carry the burden of this project.

“We have the capital from our business to do this.”

Zuckerberg.

The strategy seems to be paying off as Meta shares ticked up about 1% following the announcement, building on a gain of over 20% so far this year.

Last year, the company raked in nearly $165 billion, giving it a substantial war chest for its AI ambitions. In April, Meta raised its projected capital expenditure for 2025 to between $64 billion and $72 billion, a clear sign of its commitment to outpace rivals like OpenAI and Google.

Meta reorganized unit to offset some challenges

In response to some setbacks with its open‑source Llama 4 model and the departure of key staff, Meta recently reorganised its AI efforts under a new division called Superintelligence Labs. The unit will be headed by former Scale AI CEO Alexandr Wang and ex‑GitHub chief Nat Friedman, following Meta’s $14.3 billion investment in Scale.

Beyond raw computing power, Meta is banking on fresh revenue streams from its Meta AI app, image‑to‑video advertising tools, and upcoming smart glasses. Analyst Gil Luria of D.A. Davidson remarked that while AI has already boosted Meta’s ad sales and pricing, this scale of investment is aimed at securing the company’s lead in developing the foremost AI model, a prize that may take years to fully realise.

Zuckerberg himself has been personally spearheading an aggressive recruitment drive, scouting top engineers to join the Superintelligence Labs, with an initial target of a 50-member team, following frustrations over the company’s stalled AI progress. With heavyweights like Wang and Friedman on board, Meta hopes to cement its position at the vanguard of AI, even if the payoff is still some ways off.

As the company moves ahead with its titan clusters and new lab division, all eyes will be on whether these super‑scale data centres truly usher in a new era of machine intelligence, or if they simply raise the stakes in an already fierce technological arms race.

KEY Difference Wire helps crypto brands break through and dominate headlines fast