Meta Platforms CEO Mark Zuckerberg has announced plans to invest hundreds of billions of dollars into building some of the largest AI data centres in the world, doubling down on his vision to lead the race toward artificial superintelligence.
In a bold post on his Threads platform on Monday, Zuckerberg said the company’s first multi-gigawatt AI data centre — codenamed Prometheus — is set to come online in 2026, with another, Hyperion, designed to scale up to 5 gigawatts in the years ahead. “We’re building multiple more titan clusters as well,” he said, adding that “just one of these covers a significant part of the footprint of Manhattan.”
The announcement comes amid intensifying competition among tech giants to develop AI systems that can outperform humans on complex tasks. Meta, Google, OpenAI, Microsoft and Amazon are all racing to build massive AI infrastructure and secure top AI talent, sparking what some insiders are calling a “superintelligence arms race”.
Zuckerberg’s revelation follows a report by SemiAnalysis, which noted that Meta may be the first AI lab to bring a 1-gigawatt-plus supercluster online — a monumental leap in AI training power.
Meta’s aggressive AI push is being financed through the company’s strong advertising business, which generated nearly $165 billion in revenue in 2023. Despite investor concerns about whether such enormous spending will yield immediate returns, Zuckerberg sought to assure markets that Meta has the financial muscle to pull it off. “We have the capital from our business to do this,” he said.
Meta shares edged up by 1% following the announcement and have gained over 20% in 2025 so far.
Industry analyst Gil Luria of DA Davidson noted that Meta’s AI investment is already paying off in the short term by improving ad targeting and pricing through more powerful AI models. However, he added, “At this scale, the investment is more orientated to the long-term competition to have the leading AI model, which could take time to materialise.”
Zuckerberg recently restructured the company’s AI efforts under a new division called Superintelligence Labs, following mixed performance of its Llama 4 open-source model and the exit of key engineers. The new lab is now being led by Alexandr Wang, former CEO of Scale AI, and Nat Friedman, ex-CEO of GitHub.
Meta’s pivot comes after it invested $14.3 billion into Scale AI and raised its 2025 capital expenditure forecast to between $64 billion and $72 billion — much of it earmarked for AI infrastructure and research.
In recent weeks, Zuckerberg has personally led a high-profile hiring spree, luring elite engineers from rival companies to accelerate Meta’s AI roadmap. Reports also suggest that the company is now considering abandoning its open-source Behemoth model in favour of a closed, proprietary AI model — signalling a shift in strategy as the battle for next-generation AI dominance heats up.
Meta’s mega-investment in AI puts it in direct competition with OpenAI, Google DeepMind, and Anthropic, all of which are developing increasingly capable AI systems. However, while other companies are hedging with partnerships and paid API models, Meta appears committed to building in-house supercomputing capacity on a scale never seen before.
The company is banking on applications such as the Meta AI assistant, image-to-video ad tools, and smart glasses integrations to eventually generate revenue streams that can justify the upfront capital injection.
Click the link Puretvonline.com | WhatsApp Channel to join the WhatsApp channel
GOT A STORY?
Contact/WhatsApp: +233243201960 or Email: manuelnkansah33@gmail.com