“By the numbers:Meta said it expects 2026 capital expenditures of between $115 billion to $135 billion. That’s up from $72.22 billion in 2025.”
“For the fourth quarter, Meta’s capital expenditures were $22.14 billion, compared to expectations of $21.97 billion, CNBC reported citing StreetAccount.”
“The company said it expected higher operating income in 2026. It reported $83.3 billion in 2025, up 20% from the prior year.”
“Meta’s results increasingly have been evaluated on how aggressively it’s spending on its AI ambitions.”
“Earlier this month, Zuckerberg announced a new “top-level” initiative called Meta Compute to build “tens of gigawatts” of AI infrastructure this decade.”
So flashing green lights continue from investors on the AI investment ramps, with the stock up over 10% post the Q4 results.
“Mark Zuckerberg has expanded his direct reports to include Meta Compute’s leaders.”
“Meta Compute seeks talent for a commodities-like trading desk approach.”
“Our new org chart shows the 167 top leaders at Meta, up from 148 in our previous org chart.”
His org structure is getting clearer going into 2026:
“Meta Platforms CEO Mark Zuckerberg has expanded his direct reports to include the two co-heads of the company’s newly created Meta Compute team, signaling the importance of the new effort to manage Meta’s AI computing capacity needs.”
“Zuckerberg announced Meta Compute in mid-January, declaring it a new “top-level initiative” that will be central to its plans to build hundreds of gigawatts’ worth of AI computing capacity over time. As reflected in our updated Meta org chart, the two co-heads of Meta Compute—Santosh Janardhan, head of global infrastructure, and Daniel Gross, vice president of product—now report directly to Zuckerberg.”
The leadership structure has expanded as well, with reshaped reporting lines:
“In total, our new Meta org chart shows 167 senior leaders at Meta, up from 148 in our previous org chart. (We’ve also updated a separate org chart for Meta Superintelligence Labs.)”
“The org chart shows how the Meta Compute team is taking shape. Janardhan, a decadelong veteran of Meta who previously reported to Chief Operating Officer Javier Olivan, manages 10 direct reports, many of whom are also Meta veterans. Gross, who previously reported to Chief AI Officer Alexandr Wang, currently has no direct reports. Both Wang and Gross joined Meta last year as it launched a recruiting blitz to build up its AI talent.”
“Janardhan will continue to oversee the design, construction and operation of Meta’s global data centers, while Gross will lead a newly formed group focused on capacity strategy, industry analysis and supplier partnerships, Zuckerberg said. Both men will work closely with Dina Powell McCormick, who joined Meta as president and vice chair earlier this month, Zuckerberg said.”
And Meta has continued ambitions for more hires:
“Gross has dropped hints about his own recruiting push for Meta Compute. In a recent post on X, he said he is hiring people with backgrounds in “deep learning, supply chains, commodities, semiconductors, sovereigns, energy, Excel, prediction markets, monitoring situations, etc.”
“Gross’s recruitment post signals that Meta could run its compute division more like a commodities or trading desk at a financial institution. By recruiting expertise in commodities, energy, and prediction markets, the company appears to be positioning itself to hedge against swings in power and hardware costs and to make long-term bets on energy prices, supply chains and geopolitical constraints.”
“Meta has been pouring billions into its AI infrastructure ambitions, committing more than $600 billion in the U.S. by 2028 to scale its data center footprint. That spending shows no signs of slowing: The company in October said it expects the growth in 2026 capital expenditures will be greater, in dollar terms, than the growth in 2025. Meta projected it would spend between $70 billion and $72 billion in 2025, up from $39 billion in 2024, as it accelerated data center expansion and ramped up graphics processing unit capacity to pursue its quest for what the company calls personal superintelligence. The company reports its fourth-quarter results next Wednesday.”
“Meta’s approach contrasts with those of rivals such as Microsoft and Google, which have tightly integrated AI compute planning within their cloud businesses. Meta does not operate a cloud service, which means it cannot generate revenue by renting out cloud servers to outside customers. That raises the stakes for the company if it overbuilds or misjudges future AI demand.”
“Other new additions to Zuckerberg’s direct reports include Powell McCormick and C.J. Mahoney, Meta’s new chief legal officer and the replacement for Jennifer Newstead, who left to become Apple’s next general counsel. Former Chief Technology Officer Mike Schroepfer, now a senior fellow focused on AI, and Zuckerberg’s longtime chief of staff, Andrea Besmehn, also report to the CEO.”
Of course Meta has seen departures of some of its folks to OpenAI and beyond:
“The top ranks of Meta Superintelligence Labs have seen several significant changes since our last org chart. Ahmed Al-Dahle, former generative AI org head, joined Airbnb as CTO earlier this month, and Yann LeCun, former chief AI scientist, left to pursue a new startup. Ex-PyTorch leader Soumith Chintala joined Mira Murati’s Thinking Machines Lab in November, while Cathy Huang joined MSL from elsewhere within Meta.”
“Meanwhile, Meta Chief Strategy Officer Dave Wehner’s remit has expanded sharply since our last org chart: He now has 14 direct reports, compared with just two previously.”
“Meta has also recently begun onboarding employees from Rivos, the chip startup it said last year it was in the process of acquiring for an undisclosed sum, according to a person familiar with the matter. Bringing chip design talent in-house could complement Meta Compute’s focus on long-term planning and reduce its dependence on third-party vendors as competition for advanced chips intensifies, the person said. It also showcases Meta’s push toward greater vertical integration, potentially giving the company more control over the performance, cost and availability of the custom silicon that will power its next-generation AI workloads.”
Overall, the piece is a timely update on Meta’s frenetic top-down pace in AI investments across the board vs its peers.
Under ‘Zuck’s’ keen eye, Meta remains a ‘Founder Mode’ AI company to watch in this AI Tech Wave. Stay tuned.
(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)
Balance sheets How much cash is too much for corporations to hold? (morganstanley.com) Martin Fridson reviews “The Financial Restructuring Tool Set:...
I do NOT post twice a day on this newsletter as a habit but today I am making an exception today. When I stepped back in as CEO of Stocktwits in March of...
