Meta Enterprise Platform: Zuckerberg Turns the Social Media Giant Into an A.I. Neocloud
Trending Topics Jakob Steinschaden ● Covered by 3 sources
Meta wants to sell its AI stuff to companies, not just use it itself. That turns its data-center splurge into a new business, and puts it up against AWS, Azure and Google Cloud.
Based on reporting by Trending Topics, Jakob Steinschaden — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Meta has built a new sales pitch around the giant pile of infrastructure it has been funding for itself. The company is launching Meta Enterprise Platform, a business unit that will package its AI models, agents and infrastructure for companies and developers. For Meta, this is a first. For the market, it is a clear sign that the social media giant wants a slice of enterprise software and cloud revenue, not just ad dollars.
The unit will be run by Chirantan “CJ” Desai, who is leaving the CEO job at MongoDB to become Chief Enterprise Platform Officer and report directly to Mark Zuckerberg. Desai has also led product and engineering at Cloudflare and spent nearly eight years at ServiceNow, most recently as President and COO. Zuckerberg’s message is simple: Meta wants businesses to see it as a place to build, buy and scale AI.
The offering starts with a handful of named pieces. Muse, the personal AI agent Meta introduced for consumers a few weeks ago, is moving into companies. Meta Business Agent, which already answers customer questions, books appointments and closes sales across WhatsApp, Instagram and Messenger, is part of the bundle too. So are Muse API, which gives programmatic access to Meta’s models, and Muse Code, the coding agent that Meta says is meant to compete with Anthropic’s Claude Code and OpenAI’s Codex on aggressive pricing.
This is not a side project. Meta says it will spend between $130 billion and $145 billion on capital expenditures this year, mostly on AI data centers. So far that infrastructure has been built for Facebook, Instagram, Meta’s ad system and its own models. Now Meta wants to sell access to models, agents and eventually raw computing power as well — the same basic business that neoclouds like CoreWeave, Nebius and Lambda are chasing.
The timing tells you why this is happening. Zuckerberg has already told analysts that Meta sees a large enterprise opportunity in APIs, business agents and possibly selling compute directly, and that compute could be sold at a premium. He also admitted it would be foolish to sell all of it for short-term profit. In other words: the company has spent like a cloud provider and now wants to act like one. The only question is whether businesses outside Meta’s ad orbit will buy the pitch.
My take — AI-written commentary, not fact-checked reporting
This is classic Zuckerberg: build the expensive thing first, then figure out the invoice later. The real test is not whether Meta can offer AI to advertisers already trapped in its ecosystem; it is whether anyone outside that cosy pile will trust another giant platform with their stack. Cloud with a social-media accent is still cloud, and buyers usually notice the smell.
Read more about this at: Trending Topics