Meta Platforms, under the leadership of CEO Mark Zuckerberg, has announced an ambitious plan to invest hundreds of billions of dollars in constructing massive AI data centers to advance superintelligence, a technology aimed at surpassing human cognitive capabilities. This move intensifies Meta’s competition with tech giants like OpenAI and Google in the race to develop cutting-edge AI systems.

The initiative includes the development of multiple multi-gigawatt data centers, with the first, named Prometheus, set to become operational in 2026. Another facility, dubbed Hyperion, is designed to scale up to 5 gigawatts over the coming years. Zuckerberg emphasized the scale of these projects, noting that a single data center could cover a significant portion of Manhattan’s footprint. He also cited a SemiAnalysis report, which suggests Meta is on track to be the first AI lab to launch a gigawatt-plus supercluster, underscoring its leadership in AI infrastructure.

Meta’s financial strength, driven by its core advertising business generating nearly $165 billion in revenue last year, underpins this massive investment. Zuckerberg addressed investor concerns about the expenditure, stating, “We have the capital from our business to do this.” Meta’s stock, which has risen over 20% in 2025, saw a 1% increase following the announcement.

The company recently restructured its AI efforts under the newly formed Meta Superintelligence Labs (MSL), led by former Scale AI CEO Alexandr Wang and ex-GitHub CEO Nat Friedman. This follows Meta’s $14.3 billion investment in Scale AI and comes after challenges with its open-source Llama 4 model, including key staff departures. Reports indicate that Meta is considering abandoning its powerful open-source model, Behemoth, in favor of a closed-source alternative to strengthen its competitive edge.

Meta’s aggressive push includes a talent acquisition spree, with Zuckerberg personally recruiting top AI engineers from rivals like OpenAI, Google, and Anthropic, offering compensation packages ranging from $100 million to $300 million over four years. Notable hires include Trapit Bansal and Shengjia Zhao, both former OpenAI researchers who contributed to advanced models like ChatGPT and GPT-4. This talent war reflects Meta’s urgency to build a world-class AI team.

Analyst Gil Luria from D.A. Davidson noted that Meta’s AI investments have already boosted its advertising business by enabling more effective ad targeting and higher pricing. However, he cautioned that the scale of this investment is geared toward long-term competition for AI dominance, which may take years to yield significant returns. Meta’s increased 2025 capital expenditure, now set at $64 billion to $72 billion, underscores its commitment to closing the gap with competitors.

Meta anticipates that MSL will drive new revenue streams through products like the Meta AI app, image-to-video ad tools, and AI-powered smart glasses. With Prometheus and Hyperion leading the charge, Meta is positioning itself as a formidable player in the quest for superintelligence, though the high stakes and long timeline mean success is not guaranteed.