Meta plans cloud business to sell AI compute, taking on $200B market

Meta is building a cloud infrastructure business to sell excess AI computing power, entering a market dominated by Amazon, Microsoft and Google.

Meta's cloud push threatens to reshape the $200 billion-plus cloud computing market, pitting the social media giant against Amazon Web Services, Microsoft Azure and Google Cloud for AI compute dollars.

"It's definitely on the table," Chief Executive Officer Mark Zuckerberg told shareholders in May. "Almost every week there are different companies that come to us from the outside asking if we have compute that they could buy from us at some premium to what we've bought it at."

Meta plans to spend as much as $145 billion on capital expenditures this year building data centers and securing graphics processing units. The new business, part of an internal initiative called Meta Compute, would sell access to raw computing power or AI models hosted on Meta's infrastructure, according to people familiar with the matter. Shares of Meta jumped as much as 10% on the news.

For Meta, a cloud business offers a way to generate returns on hundreds of billions in infrastructure spending that has left investors anxious. For incumbents, it introduces a well-capitalized rival at a time when AI demand is reshaping the competitive landscape. CoreWeave, a neocloud provider that counts Meta as a customer, saw its shares fall 12% on the report.

Following SpaceX's Playbook

Meta is following a strategy pioneered by Elon Musk's SpaceX, which began selling excess computing capacity this year. SpaceX has inked deals with Anthropic for $1.25 billion per month and with Google for $920 million per month, according to Bloomberg. The strategy could help Musk's xAI generate more than $50 billion in revenue by 2028, Bloomberg Intelligence estimates.

The company is weighing two approaches: selling access to AI models hosted on its infrastructure — similar to AWS's Bedrock service — or offering raw computing capacity like neocloud providers such as CoreWeave. Meta Compute, the internal unit developing these plans, is led by Santosh Janardhan, Meta's head of infrastructure; Daniel Gross, a leader inside the Meta Superintelligence Labs AI unit; and Meta President Dina Powell McCormick.

A $145 Billion Bet on AI Infrastructure

Meta's entry comes as the cloud computing market is undergoing its biggest shift in a decade. AWS, Azure and Google Cloud have spent decades building platforms that rent access to computing power, storage and software, businesses that now generate tens of billions of dollars per quarter. The rise of AI has added a new layer: specialized chips and computing capacity for training and running models.

The move also carries risks. Building a cloud business requires not just data centers but software platforms, enterprise sales teams and customer support — areas where Meta has limited experience. The company's first model under former Scale AI executive Alexandr Wang, Muse Spark, was positioned as a "powerful foundation" rather than a state-of-the-art offering when it debuted in April.

For investors, the calculus is straightforward. Meta shares rose 10% on the news, reflecting optimism that the company can monetize its massive infrastructure bet. Nvidia, AMD and other chip suppliers could benefit from additional demand if Meta's cloud business attracts new AI workloads, though some chip stocks initially fell on concerns that increased supply could pressure pricing.

This article is for informational purposes only and does not constitute investment advice.