Meta Infrastructure Chief Santosh Janardhan to Leave in 2027

Meta's infrastructure chief Santosh Janardhan will leave the company in 2027, and former Amazon Web Services senior vice president David Brown will succeed him, CNBC reported on November 11, 2026, citing an internal memo it had obtained. Bloomberg News first reported the departure.

The timing matters. Meta just posted $34.91 billion in quarterly capital expenditures against $32.79 billion in operating cash flow, and the handover puts a new leader inside the group deciding how to divide that spending between AI development and the core business.

Who is leaving, and when?

Janardhan has worked at Meta since 2009. He currently holds two roles: head of infrastructure and co-head of engineering. His departure is not immediate. According to his internal memo, he will stay involved through the first quarter of 2027 to support the transition.

Janardhan framed the move as a planned change in leadership rather than an abrupt exit. Meta, he wrote, is making "its next set of choices about priorities," and the handover is designed to serve that moment.

The wording signals something larger than a routine personnel shuffle. Meta is mid-way through the most capital-intensive period in its history, and the person who runs the underlying compute is now changing.

Who is David Brown?

Brown joined Meta in July 2026 after serving as a senior vice president at AWS. He will formally lead Meta's infrastructure organization and take a second title: chief technical officer of Meta Compute.

Meta Compute is the unit that develops the data centers and related infrastructure behind the company's AI ambitions. Dina Powell McCormick, Meta's president and vice chairman, spearheads the initiative alongside technical leader Daniel Gross. Brown's CTO role sits within that leadership group, connecting the infrastructure organization with Meta Compute's data center and AI responsibilities.

Janardhan's memo made the case for his successor in direct terms. Brown, he wrote, has spent his career "building and operating dependable infrastructure for a wide range of demanding workloads." He described Brown as having "the instincts, depth and judgment this moment requires," and said the handover would let Brown shape Meta's next phase of infrastructure work.

The endorsement carries weight given the stakes. The infrastructure Brown now runs is the physical substrate for every AI model Meta trains and deploys.

What do the capex numbers show?

The financial backdrop comes from Meta's third-quarter 2026 results, released October 21, 2026. The numbers explain why this leadership change matters now:

• $34.91 billion in quarterly capital expenditures, including principal payments on finance leases
• $32.79 billion in cash flow from operating activities for the same quarter
• Negative $2.12 billion in free cash flow under Meta's own definition, which subtracts both property and equipment purchases and finance lease principal payments from operating cash flow

The negative free cash flow figure is the headline. It means Meta's investment outlays exceeded the quarter's operating cash generation. A company of Meta's scale is now spending more on infrastructure than its operations bring in, quarter by quarter.

For the full year, Meta narrowed its expected 2026 capital expenditures to a range of $135 billion to $145 billion, tightening an earlier outlook of $130 billion to $145 billion. The revision raised the bottom of the range by $5 billion without changing the top.

That floor increase is the telling detail. Meta did not just confirm its spending plans; it ruled out the cheaper end of them.

Why is this happening now?

Meta said it is still developing its 2027 budget and capacity plans, but the company expects to keep investing aggressively in its core business and its emerging AI initiatives. Current and scheduled capacity is intended to provide flexibility across three demands, according to the company:

• Model training
• Core-business growth
• Development and support of new products

Brown will help set those allocations as the next budget takes shape, with Janardhan providing continuity during the transition period. In effect, the incoming CTO inherits a seat at the table where Meta decides how to split scarce compute between the advertising machine that funds everything and the AI bets that could define the next decade.

The structural detail is worth noting. By placing Brown's CTO role inside the Meta Compute leadership group alongside Powell McCormick and Gross, Meta is tying its infrastructure organization directly to the initiative responsible for AI data center buildout. The move consolidates decision-making over compute at the exact moment compute decisions carry nine-figure consequences each quarter.

What comes next?

The transition runs through the first quarter of 2027, the same window in which Meta will finalize its 2027 budget and capacity plans. Brown's first major task will be helping decide how aggressively to keep spending once the numbers are locked, while the company's capital outlays exceed its operating cash generation.

For Meta, the bet is that the infrastructure Brown built his career on at AWS becomes the foundation of its AI future — and that the $135 billion to $145 billion it is spending this year buys a durable advantage rather than a balance-sheet strain.