Microsoft’s latest AI reorganization comes into view through Satya Nadella’s new company email, and it underscores how young Microsoft’s AI business still is. For all the sweeping claims about AI reshaping the company, the unit itself has only existed in its current form for a few years. That makes a leadership reset at this stage feel less like a routine tune-up and more like a sign that the business model is still settling. When a division is this early in its life and already shifting its structure again, it suggests the technology may be maturing faster than the organization built to support it.
Nadella’s update brings Microsoft’s consumer and commercial Copilot work under one umbrella, a move he frames as a step toward building a unified system rather than a collection of disconnected features. The person now responsible for that unification is Jacob Andreou, who joined Microsoft in 2023 after nearly a decade at Snap. Andreou built his reputation as the executive who helped scale Snapchat’s product and growth engines during the company’s most turbulent years. His background is rooted in consumer behavior and rapid iteration, which makes him an interesting fit for a product like Copilot that has grown quickly but not always coherently. Nadella’s email positions him as the person who can bring order to a portfolio that has expanded faster than the company’s ability to standardize it.
The other major figure in this reorg is Mustafa Suleyman, who will continue leading Microsoft’s frontier model and superintelligence efforts. Suleyman co founded DeepMind before joining Google and later co founding Inflection AI, which Microsoft acquired in 2024. His career has been defined by pushing the boundaries of model development, often faster than the companies around him were prepared to operationalize. Nadella’s message emphasizes that progress at the model layer remains essential to Microsoft’s long term AI strategy. Suleyman’s own note reinforces that point, describing frontier models as the foundation for everything Microsoft hopes to build and highlighting the company’s expanded compute roadmap. His presence signals that Microsoft still sees the model layer as the real engine of its AI ambitions, even if the organizational structure around it keeps shifting.
Other executives named in the new Copilot Leadership Team bring their own histories into the mix. Charles Lamanna has spent years inside Microsoft’s business applications and low code platforms, often serving as the bridge between enterprise customers and Microsoft’s more experimental technologies. Perry Clarke is a long time engineering leader whose career spans Exchange, Outlook, and the early days of Microsoft’s cloud transition. Ryan Roslansky, formerly LinkedIn’s CEO, brings a background in large scale professional networks and data driven product ecosystems. Together, they represent a blend of consumer, enterprise, and infrastructure experience, which suggests Microsoft is still trying to figure out which of those instincts should define Copilot’s future.
Because Microsoft’s AI business is still early in its development, the timing of this reorganization stands out. The company has positioned AI as a defining pillar of its future, yet the structure supporting that pillar continues to shift. That fluidity reflects the pace of the field, but it also hints at a business still trying to prove its long term value. AI may be the most hyped technology in the company’s history, but hype does not automatically translate into a stable operating model. When a unit is this young and already on its second or third leadership reset, it is hard not to read it as a sign that Microsoft is still figuring out what an AI business should look like at scale.
Nadella says the updated structure is designed to match the architecture of the systems Microsoft is building, while Suleyman describes it as a way to ensure that models and products reinforce each other. Both messages aim to project stability, even as the organization continues to evolve. The reorg may ultimately help Microsoft sharpen its AI strategy, but it also serves as a reminder that the company is still building the scaffolding around a technology it insists will define its future.
