Satya Nadella tells Microsoft’s top leaders to commit to AI or…

This mindset is already reshaping how Microsoft is run. Nadella has recently restructured the leadership team to free up more of his own time for technical work. Judson Althoff was promoted to chief executive of Microsoft’s commercial business, a move explicitly designed to allow Nadella to focus on datacentre expansion, systems architecture and AI-driven product development. The shift was visible at Microsoft’s recent Ignite conference, where Althoff delivered the keynote instead of Nadella, a first during Nadella’s tenure as CEO.

Internally, Nadella has also changed how innovation is driven. Reports suggest he has launched weekly AI accelerator meetings that prioritise participation from hands-on technical staff rather than senior executives. The aim is to create a deliberately messy and fast-moving environment that avoids traditional top-down decision-making. In internal communications, Nadella has urged senior leaders to behave more like individual contributors, spending less time managing and more time building, testing and learning.

In one internal message, Nadella dismissed the idea that agility only exists at start-ups. He argued that the same intense, experimental work is already happening inside Microsoft, often unnoticed, and that leaders need to immerse themselves in it rather than admire it from a distance.

As this cultural reset accelerates, further leadership changes appear likely. Longtime Office and Windows head Rajesh Jha is said to be weighing retirement, although the renewed focus on AI may yet influence his decision. If he were to step down, internal speculation has pointed to LinkedIn chief executive Ryan Roslansky as a potential successor. Roslansky’s responsibilities have already expanded to include major Microsoft 365 products and Copilot, placing him closer to the centre of the company’s AI strategy.

The transformation is also redefining how Microsoft thinks about software creation itself. Senior product leaders have described a new production model where AI systems and agents act as scalable units of output. In this framework, progress is no longer tied directly to the number of engineers or hours worked. Intelligence, not headcount, becomes the multiplier.

Underlying all of this is Nadella’s fear of irrelevance. He has spoken openly about being haunted by the collapse of once-dominant technology companies that failed to adapt. In internal forums, he has warned that even Microsoft’s most successful businesses cannot be taken for granted in an AI-driven future.

That anxiety helps explain the human cost of the transition. Microsoft laid off around 6,000 employees in May and another 9,000 in July. Nadella described the cuts as part of a painful but necessary process of unlearning old ways of working. Reports from within the company suggest morale has taken a hit, with employees describing a colder, more rigid environment than in previous years.