Zuckerberg eyes selling Meta's internal AI tools and compute to large enterprises

Meta’s enterprise AI ambitions go far beyond the customer service chatbots it launched in June. On Wednesday’s Q2 earnings call, CEO Mark Zuckerberg sketched a future where the company sells its own internal productivity and coding tools, raw compute power, and business agents to large customers — not just the millions of small business advertisers currently using its platforms.

“We see a large enterprise opportunity to sell to businesses, including APIs, business agents, potentially selling compute directly, and other services that we’re building for large customers,” Zuckerberg told investors. This represents a fundamental bet on diversifying Meta beyond advertising. The initial focus stays close to home: AI agents that work across WhatsApp, Messenger, and Instagram so businesses can handle customer interactions through an AI interface. The monetization model borrows from Meta’s ad playbook. “Just like the ad system, effectively, we will get paid when we deliver results for those businesses,” he said.

But selling to Fortune 500 companies is a different game than serving millions of small advertisers. Zuckerberg acknowledged this shift requires flexing a “different muscle” than the one Meta has historically built. The company didn’t hide the fact that its newly built internal tools — born from necessity — could now become a product line for outsiders. The play is to take developer and productivity tools tuned for Meta’s own massive scale and offer them to other enterprises.

The compute piece is especially interesting. Meta says it can sell compute at “a significant premium over what we paid for it.” But Zuckerberg isn’t racing to cash in. He called it “foolish” to sell off all that capacity for short-term profit, framing the strategy as a portfolio balancing long-term infrastructure needs against near-term revenue. He tied this to the hardware demands of “personal superintelligence” — a concept he’s increasingly using to describe Meta’s AI endgame.

The call also touched on agentic AI for consumers, from personal assistants to smart glasses, and teased a faster pipeline for shipping experimental social apps using large language models. The throughline is clear: Meta is weaving AI into every surface it owns and now, increasingly, into surfaces it wants to sell.