Meta is swapping leaders at the top of its most critical division: AI infrastructure. Longtime chief Santosh Janardhan is out, and former AWS heavyweight David Brown is in. Here’s what this strategic change signals about Meta’s AI ambitions.
The Departing Veteran
Santosh Janardhan,
Meta's head of infrastructure and co-head of engineering, is stepping down after a tenure of over 15 years. Having joined the company in 2009 when it was still just Facebook, Janardhan was instrumental in building the colossal technical foundation that supports billions of users across its family of apps. In an internal memo, Janardhan stated it was 'time to pass the baton', without specifying reasons for his departure. He will remain with the company through the first quarter of 2027 to ensure a smooth transition, after which he will move into an advisory role for Meta Compute, the company's initiative to scale its data centre investments. His departure marks the end of an era for one of the key architects of Meta's global scale.
The New Head of AI Infrastructure
Stepping into this crucial role is David Brown, an executive Meta strategically hired from Amazon Web Services (AWS) in July 2026. Brown is not just another executive; he is a cloud and compute veteran who spent nearly 19 years at Amazon. He was a pivotal figure in the early days of AWS's foundational EC2 service, which allows companies to rent computing power, and rose to become a senior vice president overseeing compute and machine learning services. His experience at AWS involved building and operating infrastructure at a planetary scale, including the development of custom silicon like Graviton and Trainium AI chips. In an internal memo, Janardhan noted that Brown possesses the 'instincts, depth and judgment this moment requires'.
Why This Change, and Why Now?
This leadership transition is happening at a pivotal moment. Meta is in the midst of an unprecedented spending spree to build out its AI infrastructure, positioning itself to compete with Google, Microsoft, and OpenAI. The company has projected capital expenditures for 2026 to be between $135 billion and $145 billion, a massive increase from previous years, with the bulk of it aimed at data centres, custom chips, and the raw computing power needed to train and run advanced AI models. This has put a strain on the company's finances, even leading to negative free cash flow in the third quarter. Bringing in an executive like Brown, who has deep experience in operating massive, efficient, and cost-effective cloud infrastructure at AWS, signals a strategic shift. The focus may be moving from simply building capacity to optimising it for performance and financial return, a discipline Brown honed at Amazon.
What This Signals for Meta’s Future
Hiring Brown and placing him at the helm of infrastructure, including as the new chief technical officer for the Meta Compute initiative, is a powerful statement. It suggests Meta is not only building infrastructure for its own AI products, like the Muse series of models, but is also seriously contemplating a future where it might rent its vast computing power to other companies, much like AWS does. Brown’s mandate reportedly includes exploring 'Meta Compute' as a potential cloud service. His expertise is in creating the 'illusion of infinite capacity' for customers, a core tenet of cloud computing. This move suggests Meta's long-term strategy involves turning its biggest cost centre—infrastructure—into a potential revenue generator, fundamentally altering its business model and creating a new front in the cloud wars.















