Meta targets cloud market, challenging Amazon, Microsoft, and Google

Meta is building a cloud infrastructure business. The company that made its fortune selling ads between photos of your cousin’s vacation is now coming for Amazon, Microsoft, and Google’s most profitable turf.

The move, reported by Bloomberg on July 1, 2026, would see Meta sell its excess AI computing capacity to outside customers. The company has capital expenditure guidance for 2026 set between $115 billion and $135 billion.

From social network to server farm

The seeds of this pivot were planted publicly at Meta’s annual shareholder meeting on May 27, 2026, when CEO Mark Zuckerberg said a cloud offering was “definitely on the table.”

Meta has been building out AI infrastructure at a pace that would make most sovereign nations blush. That $115 billion to $135 billion capex range for 2026 alone is primarily aimed at AI capacity. Rather than let those expensive GPUs collect digital dust, Meta plans to lease that surplus capacity to outside businesses. The company reportedly receives weekly external inquiries about purchasing compute capacity, despite currently utilizing all available resources.

The initial offering will focus specifically on AI-centric compute rather than attempting to build a full general-purpose cloud stack, with Meta playing to its strengths in AI model development, including its Llama model family.

The irony of Meta’s cloud partnerships

Meta signed a $10 billion, six-year deal with Google Cloud back in August 2025. It also expanded a partnership with GPU cloud provider CoreWeave worth $21 billion, extending through December 2032. Combined, that’s $31 billion in commitments to other people’s clouds.

Can a social media company win enterprise trust?

The cloud infrastructure market isn’t just about having servers. It’s about trust. Enterprises choosing a cloud provider are making a decision that affects their data security, regulatory compliance, and operational reliability for years. Meta’s track record on data handling is, to put it charitably, complicated. The company has faced years of scrutiny over privacy practices in its social media business.

That said, Meta’s decision to focus narrowly on AI compute rather than general-purpose cloud services partially sidesteps this concern. Companies looking to train or run AI models may care more about raw performance and cost than about whether their cloud provider also runs Instagram.

Investors should watch for concrete timelines on Meta’s first external cloud customers, any enterprise partnerships that signal market acceptance, and whether the company begins breaking out cloud revenue as a separate line item in earnings reports.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.