Microsoft appoints a new Copilot boss after AI leadership shake-up

```html

Microsoft is restructuring again, but this time it's about something more serious than just a change of positions. After more than three years of experimenting with separate versions of Copilot for consumers and business, the Redmond giant finally admits it was wrong — that strategy didn't work. The new head of Copilot, Jacob Andreou, has the task of bringing together what should have been together from the beginning, while Mustafa Suleyman, the former head of AI, moves to a more abstract role of creating Microsoft's own artificial intelligence models. This isn't just a reshuffling of positions — it's a signal that Microsoft finally realizes the chaos it created around its flagship AI assistant.

Chaos at home, or how Copilot became two different products

If you've ever used Copilot on your computer and then switched to the business version at work, you probably felt like you were moving between two completely different applications. It wasn't an impression — it was reality. For years consumer and corporate teams worked independently, creating assistants that had almost nothing in common. It's not just about the look of the interface, but about fundamental differences in functionality, capabilities, and even the approach to user interaction.

When Mustafa Suleyman joined Microsoft almost two years ago, he brought with him a team from Inflection AI — a startup that worked on the Pi chatbot. Shortly after, the consumer version of Copilot underwent a major redesign that looked suspiciously similar to what Inflection had done before. Meanwhile, the business version of Copilot remained in its separate world, integrating with Microsoft 365 and focusing on corporate productivity. It was as if Microsoft had two different visions of what Copilot should be, and never managed to reconcile them.

The result? Confused users, internal teams working at cross purposes, and Copilot losing brand coherence. Nobody really knew who was responsible for the overall strategy of the assistant — both within Microsoft and in the eyes of the public.

Suleyman moves to abstraction, Andreou takes on the concrete

The new structure is clear: Jacob Andreou will be responsible for the entire Copilot user experience — from design to engineering, for both consumers and business. This means that finally one person has the authority to introduce coherence. Andreou is not a new name in the industry — he previously worked at Snap in positions related to product and growth, and joined Microsoft AI just a year ago. Now he reports directly to Satya Nadella, which gives him a mandate to act.

Meanwhile, Mustafa Suleyman, instead of struggling with the day-to-day challenges of Copilots for different markets, is focusing on creating Microsoft's own AI models. This is a move that makes sense — Suleyman has experience building AI foundations, and Microsoft really does need models tailored to its specific business needs. Suleyman will maintain some control over day-to-day operations through a "dotted line" to Andreou, but his main focus will now be on models, not interfaces.

In practice, this means Suleyman is stepping out of the front lines. Instead of being the one who has to explain why Copilot on your laptop looks different than at work, he can now focus on something that has always interested him more — AI engineering at a fundamental level.

New leadership team and four pillars of Copilot

Above Andreou is a team of four key leaders: Ryan Roslansky, Perry Clarke and Charles Lamanna. Each of them has clearly defined responsibilities within four pillars that Nadella mentioned in his internal memo:

• Copilot experience — user experience, interface, interactions
• Copilot platform — infrastructure, APIs, integrations
• Microsoft 365 apps — integration with office applications
• AI models — artificial intelligence models powering everything

This is a structure that makes sense. Instead of chaos with separate teams working in parallel without coordination, we now have a hierarchy with clear information flow. Roslansky, Clarke and Lamanna will work on Microsoft 365 applications and the Copilot platform, while Andreou will coordinate the entire user experience across all platforms.

Nadella in his memo talks about transitioning "from a collection of great products to a truly integrated system". This sounds like an admission that the previous approach didn't work. But it's also a vision — if they can pull it off, Copilot could become something competitors (OpenAI, Google, Anthropic) haven't yet achieved: a truly integrated assistant that works as well for a startup as it does for Fortune 500.

Questions about the future of Bing, Edge, and advertising

Here's a question that The Verge rightly raises: what will happen to Microsoft Edge, Bing, MSN and the advertising business? All these teams have so far reported to Suleyman. Now that Suleyman is moving to AI models, these products are left without a leader. Microsoft didn't explain this in the announcement, which suggests that further changes may be on the way.

This is particularly important for Bing. Three years ago, Microsoft made a huge fuss about Bing Chat — a revolutionary AI assistant integrated with the search engine. It was a response to ChatGPT, an ambitious attempt to redefine search. Then Microsoft rebranded Bing Chat to Copilot and moved it to consumer applications. Now Bing sits in limbo — will it be an independent product? Will it be completely absorbed by Copilot? Or has Microsoft created a problem for itself that it can't solve?

For Polish users, this means we can expect changes in how Bing and Edge function with Copilots over the coming months. Perhaps it will finally be a coherent experience instead of a weird set of unrelated features.

Series of retirements and changes at the top

This change doesn't happen in a vacuum. Just a week earlier, Rajesh Jha, who worked at Microsoft for over 35 years and oversaw Microsoft 365 Copilot, Windows, Office and much more, announced his retirement. This is a significant loss — Jha was one of the architects of how Copilot integrates with Microsoft's application ecosystem.

Last month, Phil Spencer, the legendary head of Xbox, also announced his retirement after almost 40 years at the company. Asha Sharma took his place. These departures indicate a deep reorganization of Microsoft — this is not just a change of positions, it's a change of leadership generations.

For the AI industry, this is a signal that even giants must rebuild when their strategies don't work. Microsoft invested heavily in Copilot, but the organizational structure wasn't able to support the product vision. Now they're trying to fix it.

Deep causes — why Copilot fell apart

To understand why this change was necessary, you have to go back to the beginning. Microsoft had two different visions for Copilot: one for consumers, which was to be more personal, conversational, inspired by Inflection AI's work on Pi, and another for business, which was to be a productivity tool integrated with Microsoft 365.

The problem was that these two visions never met. Teams worked independently, with different priorities, different metrics of success, and different interpretations of what Copilot should be. For consumers, it was an assistant that was supposed to be friendly, interactive, almost like talking to a friend. For business, it was a tool that was supposed to increase productivity, write emails, create documents.

Reality? Users were confused. When they switched between versions, they felt like they were operating two completely different products. This weakened the Copilot brand and made it difficult to explain what its real potential was.

What this means for Copilot's future

The new structure suggests that Microsoft wants to create one Copilot that works everywhere — on your laptop, phone, at work, at home. It's ambitious, but logical. If OpenAI can make ChatGPT work for everyone, why couldn't Microsoft do something similar, but better integrated with the Windows and Microsoft 365 ecosystem?

Andreou has experience in product growth, which suggests that Microsoft not only wants to unify Copilot, but also promote it. This could mean more aggressive marketing, better integration with Windows, maybe even free consumer plans that are more attractive.

The question is: will they be able to do it better than before? Suleyman had good intentions, but perhaps he was too busy building AI models to focus on user experience coherence. Andreou, with his experience at Snap, might be exactly who is needed to make Copilot a product that people actually want to use.

For the Polish market, this means we can expect greater coherence between Copilots in different Microsoft products. If Andreou does his job well, Copilot could finally become something that makes sense for regular users, not just tech enthusiasts.

```