Meta CEO Mark Zuckerberg holds a smartphone while delivering a keynote speech at the annual Meta Connect event at the company’s headquarters in Menlo Park, California, USA, on September 25, 2024.
Manuel Orbegozo | Reuters
As Meta Deciding whether to sell or retain excess capacity is a balancing act, according to CEO Mark Zuckerberg, who is gobbling up land to build massive AI data centers.
Among the big four U.S. hyperscalers, Meta is the only one not in the business of selling cloud infrastructure and services, although its capital spending rivals that of its peers. But in recent months, Zuckerberg has floated the possibility of launching a cloud business to leverage his spare capacity in a resource-constrained market.
As CNBC reported earlier this month, Anthropic is in preliminary talks to lease computing power from Meta.
“We’re getting a lot of offers for computing power at a significant premium to what we paid for it,” Zuckerberg said on the company’s second-quarter earnings call after the bell on Wednesday. “And we also have more programming and productivity tools on our roadmap.”
Zuckerberg’s comments came after Meta gave weaker-than-expected third-quarter revenue guidance and said free cash flow fell 90% from a year ago due to rising capital spending. The stock fell more than 7% in after-hours trading, extending a slump that pushed it down 11% for the year through Wednesday’s close.
In the report, Meta raised the low end of its 2026 investment forecast by $5 billion, bringing the range to $130 billion to $145 billion. Last week, alphabet raised the high end of its forecast to $205 billion and turned cash flow negative for the first time. And Microsoft said in its earnings report Wednesday that capital spending for the year will be about $175 billion. Amazon reports results on Thursday.
Investors were looking for details on Zuckerberg’s AI strategy, but it was scattered, leaving Meta behind OpenAI, Anthropic and Google in the market for top models and services.
“I think everyone wants clarity about what they want to do in the computer business,” Jefferies analyst Brent Thill told CNBC’s “Closing Bell Overtime.”
Table of Contents
Considering the compromise
Zuckerberg offered little specifics about his plans, but did explain some of the various considerations at play.
“In terms of running the company, of course we have to make a common compromise: How much do you monetize something today and develop future assets?,” Zuckerberg said. “I think it’s always a portfolio.”
He said that when looking at a potential corporate deal, it’s not just about selling capacity. The company also has API and productivity services it can offer, as well as AI agents it is developing, Zuckerberg said.
“And I think there’s a very, very big opportunity there,” he said.
However, Meta needs sufficient computing capacity to fulfill its own AI ambitions, especially as the company begins aggressively introducing new models under the leadership of AI chief Alexandr Wang. Earlier this month, Meta introduced the Muse Spark 1.1 model, which Wang said is the “strongest model for agent and coding work to date” and is available at a cheaper price than OpenAI and Anthropic’s offerings.
Meta Chief AI Officer Alexandr Wang during the Bloomberg Tech conference in San Francisco, California, USA, on Thursday, June 4, 2026.
David Paul Morris | Bloomberg | Getty Images
“It would be stupid to basically just sell all the computing power and take a short-term profit,” Zuckerberg said.
Zuckerberg acknowledged that joining the company, where Meta has struggled in the past, will take some hard work and that the company will have to learn how to do it. Although he didn’t mention hiring a sales force, this will be an essential step if Meta is serious about selling to companies large and small.
“This will be kind of a new muscle that we build as a company,” Zuckerberg said. “But I think it’s a very important project that we’re building.”
Dave Brown, a former longtime executive at Amazon Web Services, is set to join the company, CNBC recently confirmed.
A major challenge for Zuckerberg as he tries to convince Wall Street of his vision is his patchy track record. Zuckerberg’s most notable whiff has been the metaverse and his attempt, beginning in 2021, to reshape the company around a futuristic digital world.
This project still costs Meta billions of dollars per quarter. Metas Reality Labs, which develops virtual reality devices and wearables, lost $4.62 billion on revenue of just $431 million in the most recent period.
Still, Meta is desperate to diversify its business beyond digital ads, which still account for 98% of the company’s revenue, and show that it can be an influential player in AI.
Zuckerberg is all in.
“I understand that this is kind of a big bet across the industry,” Zuckerberg said. “My personal bet is that people who invest here will be rewarded and feel very good over time.”
REGARD: Brent Thill from Jefferies: The meta report was a “threefold” of things that could go wrong.
Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.
https://www.cnbc.com/2026/07/29/zuckerberg-metas-ai-capacity-dilemma-what-to-sell-vs-what-to-keep.html
