Meta Platforms shares jumped more than 6% in premarket trading on Wednesday after Bloomberg News reported the company is building a cloud business to sell excess AI computing capacity to external customers. The move, citing people familiar with the matter, would transform Meta from a pure consumer-social platform into a third-party infrastructure provider – putting it in direct competition with Amazon Web Services, Microsoft Azure, and Google Cloud. YourDailyAnalysis flags the market reaction as the real signal: a 6% premarket move on a single Bloomberg report, before any official confirmation from Meta, shows just how aggressively investors are pricing in the AI infrastructure upside that has already propelled rivals’ cloud revenues.
Start with the strategic context. Meta has spent the last two years committing to AI infrastructure at a scale that the company’s own internal workloads cannot fully absorb. CEO Mark Zuckerberg announced in January 2025 that Meta would spend up to $65 billion on capital expenditures that year, with a large share directed at AI compute buildout. The company’s AI models – including Llama, its open-source large language model family – run on proprietary infrastructure that now represents one of the largest private GPU clusters in the world. The logical extension of over-building capacity is exactly what Bloomberg described: monetise the excess by selling compute time. Nvidia has benefited enormously from Meta’s purchasing, and the question now is whether Meta can capture a portion of the margin that currently flows to cloud providers selling that same compute to others.
The competitive read matters. Amazon Web Services generated around $107 billion in revenue in 2024 and has consistently grown at double-digit rates driven by enterprise AI workloads. Microsoft Azure crossed $100 billion in annualised revenue partly on the strength of its OpenAI partnership. Google Cloud hit $43 billion in 2024 revenue. Meta entering this market as a seller rather than a buyer would alter the supply dynamics for AI compute access, particularly for smaller companies and startups that currently rely on third-party cloud providers for GPU time. There is a counter-argument that YourDailyAnalysis surfaces: Meta’s existing cloud infrastructure is optimised for its own recommendation and ad-targeting workloads, which may not translate cleanly into general-purpose AI compute that enterprise customers expect from a hyperscaler.
The financial logic is compelling even at small scale. If Meta can monetise even 5-10% of its installed compute capacity at market rates for GPU-hour pricing, the revenue contribution would be material against a cost base that is already largely fixed. The company does not disclose granular infrastructure utilisation rates, but industry observers have estimated that large-scale AI operators typically run clusters at 60-80% utilisation at peak, with significant headroom at off-peak periods. Editors at YourDailyAnalysis weigh this headroom as the real commercial opportunity: selling the idle portion of a fixed-cost asset at near-zero marginal cost is a genuinely high-margin business if demand can be matched to supply.
There is a structural question about what Meta would actually be selling. Hyperscalers offer not just raw compute but a full ecosystem of storage, networking, developer tooling, and compliance frameworks that enterprise customers depend on. A bare-metal GPU rental business is a very different product from what AWS or Azure provides, and the enterprise sales motion – long procurement cycles, security reviews, contractual commitments – is foreign to Meta’s existing business development culture. The cleanest version of this story is that Meta starts with AI companies and research institutions that need raw GPU access without the enterprise overhead, and builds outward from there. Less clean is any version that requires Meta to compete head-on with hyperscalers on the full stack.
The timing of the leak is not random. Meta reports quarterly earnings in late July, and a report framing the company as an emerging cloud infrastructure player lands at the precise moment when analysts are building their models for the second half of the year. Reporters at Your Daily Analysis note that the stock was already up roughly 40% year-to-date before Wednesday’s premarket move, reflecting a market that has been repricing Meta as an AI infrastructure story rather than a social media platform. Watch the Q2 earnings call for any executive commentary that either confirms or distances from the cloud business report – that verbal signal will set the investment thesis for the rest of 2026 more durably than any premarket move on an unconfirmed Bloomberg item.
