What to sell vs. what to keep

As Meta gobbles up land to construct massive AI data centers, CEO Mark Zuckerberg says there’s a balancing act when it comes to deciding whether to sell excess capacity or to preserve it.

Among the four major U.S. hyperscalers, Meta is the only one that doesn’t have a business selling cloud infrastructure and services even though its capital expenditures rival that of its peers. But in recent months, Zuckerberg has been floating the possibility of launching a cloud business as a way to take advantage of its capacity stockpile 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 compute at a significant premium over what we paid for it,” Zuckerberg said on the company’s second-quarter earnings call after the bell on Wednesday. “And we have more coding and productivity tools on our roadmap as well.”

Zuckerberg’s comments came after Meta issued a weaker-than-expected revenue forecast for the third quarter and said free cash flow dwindled by 90% from a year earlier due to soaring capex. The stock sank more than 7% in after-hours trading, extending a slump that’s pushed it down 11% for the year as of Wednesday’s close.

In the report, Meta bumped up the low end of its 2026 capital expenditure guidance by $5 billion, bringing the range to between $130 billion and $145 billion. Last week, Alphabet hiked the top end of its guidance to $205 billion and turned cash flow negative for the first time. And Microsoft said in its earnings report on Wednesday that capex for the year will be roughly $175 billion. Amazon reports results on Thursday.

Investors have been seeking details on Zuckerberg’s AI strategy, which has been scattershot and left Meta behind OpenAI, Anthropic and Google in the market for top models and services.

“I think everyone wants clarity into what he wants to do in the compute business,” Brent Thill, an analyst at Jefferies, told CNBC’s “Closing Bell Overtime.”

Zuckerberg offered little by way of specifics on his plans, but he spelled out some of the various considerations at play.

“In terms of running the business, obviously, a common trade-off that we need to make is around how much do you monetize something today versus develop future assets,” Zuckerberg said. “I think that it’s always a portfolio.”

He said that in looking at a potential enterprise business, it’s not just about selling capacity. The company also has API and productivity services it can offer as well as AI agents that it’s building, Zuckerberg said.

“And I think that there’s just a very, very large opportunity there,” he said.

However, Meta needs ample compute capacity to satisfy its own AI ambitions, especially as the company begins aggressively rolling out new models under the leadership of AI chief Alexandr Wang. Earlier this month, Meta debuted the Muse Spark 1.1 model, which Wang said represents the “strongest model for agentic and coding work yet” and at a cheaper price than offerings from OpenAI and Anthropic.

“It would be foolish to basically just sell all of the compute and take a short-term profit,” Zuckerberg said.

Zuckerberg acknowledged that jumping into the enterprise, where Meta has historically struggled, will require some hard work, and that the company has to learn how to do it. While he didn’t reference hiring a sales force, it will be an essential move if Meta is serious about selling to businesses big and small.

“That’s going to be somewhat a new muscle that we build as a company,” Zuckerberg said. “But I think it’s a very important one that we build.”

Dave Brown, a former longtime senior executive at Amazon Web Services, is set to join the company, CNBC recently confirmed.

A big challenge for Zuckerberg as he tries to sell Wall Street on his vision is his spotty track record. Zuckerberg’s most notable whiff was the metaverse, and his effort starting in 2021 to reshape the company around a futuristic digital world.

That project is still costing Meta billions of dollars a quarter. Meta’s Reality Labs, which develops virtual reality devices and wearables, lost $4.62 billion in the latest period on just $431 million of revenue.

Still, Meta is desperate to diversify its business beyond digital ads, which still account for 98% of the company’s revenue, and to show that it can be an influential player in AI.

Zuckerberg is all-in.

“I get that this is sort of a big bet across the industry,” Zuckerberg said. “My personal bet is that the people who invest in this are going to be rewarded and feel very good over time.”

WATCH: Jefferies’ Brent Thill: Meta report was ‘trifecta’ of things that could go wrong.