Microsoft merges its consumer and enterprise Copilot teams to build a single platform

Microsoft is merging the teams that develop consumer and enterprise Copilot. The company wants to remove the internal boundaries between the two versions of the assistant, ship features faster and make the product more cohesive for users at home and at work.

Why change the structure

Until now, Copilot inside Microsoft lived in several parallel tracks: separate teams were responsible for the consumer experience, enterprise scenarios, Microsoft 365 apps and platform components. That approach helped Microsoft roll out AI across its ecosystem quickly, but over time it created fragmentation. One brand covered different interfaces, different capabilities and different logic, from Windows and web chat to Word, Excel, Outlook and Teams.

Now Microsoft wants to turn Copilot from a set of linked products into a single system. Internally, the company ties this directly to a new stage of AI: assistants should not only answer questions but also carry out multi-step tasks, coordinate actions across apps and work as an agent. For that scenario, shipping features one by one is no longer enough. It takes a shared architecture, common product priorities and fewer internal gaps between work and personal use cases.

Who is in charge now

Microsoft is building the unified Copilot around four areas: the Copilot user experience, the Copilot platform, Microsoft 365 apps and AI models. Jacob Andreou, who becomes executive vice president for Copilot and reports directly to Satya Nadella, now owns the shared experience for consumer and enterprise users. Ryan Roslansky, Perry Clark and Charles Lamanna will continue to run the apps and the platform layer, and are expected to tie Copilot to Microsoft 365 and the company's infrastructure.

"We are moving from a set of great products to a truly integrated system," is how

Satya Nadella described the point of the overhaul in a message to employees.

In practice, this means Microsoft wants to end the situation in which the same Copilot feels like several different products under one sign. Users should see a more consistent interface, and the team should be able to move successful features faster from the enterprise side to the consumer side and back. For the Copilot brand this is an important step: the company is betting not on a scatter of separate AI entry points but on one main assistant layer on top of its services.

A bet on models and agents

A separate part of the reorganization concerns Mustafa Suleyman. He is shifting his focus from day-to-day product operations to the model layer and Microsoft's long-term AI strategy. In a message to employees, Suleyman explained that over the next five years the company wants to spend its effort building its own world-class models that fit enterprise tasks better, creating separate model lines for business scenarios and lowering the cost of serving large AI workloads. For Microsoft this is critical if Copilot is to scale to millions of users and increasingly autonomous agents.

The reorganization comes right after a series of major announcements in March. Microsoft has already introduced Wave 3 for Microsoft 365 Copilot, the Agent 365 service and Frontier Suite for enterprise customers, and is separately promoting Copilot Cowork and Work IQ. According to the company itself, paid Copilot seats grew by more than 160% year over year, daily usage increased tenfold, and 90% of Fortune 500 companies already use the product. This is the point at which scattered development starts to get in the way of scale.

• Wave 3 adds new agentic scenarios to Word, Excel, PowerPoint and Outlook
• Agent 365 acts as a layer for managing, monitoring and securing AI agents in a company
• Copilot Cowork is built for long, multi-step tasks that are not completed in a single request
• Work IQ gives Copilot context from work data, documents and relationships inside an organization

All of these pieces require a tighter link between the interface, the platform, the apps and the models themselves. That is why merging the teams looks less like a cosmetic reshuffle and more like an attempt to prepare Copilot for the next phase of competition, where what matters is not only AI demos but real usefulness, manageability and price for mass use. For Microsoft it is also a way to turn AI research ideas into a product that is equally clear to a company employee and an ordinary user, faster.

What this means

Microsoft is moving Copilot from a collection of scattered features to a single AI platform for work and personal tasks. If the company really does pull the product into one coherent experience and strengthens its own model layer at the same time, Copilot has a better chance of becoming not just AI built into the Microsoft ecosystem but a central interface for everyday work, both at home and in the office.